PJFP.com

Pursuit of Joy, Fulfillment, and Purpose

Tag: Project Glasswing

  • Claude Opus 4.8 Released: Anthropic Bets on Honesty, Dynamic Workflows, Effort Control, and Cheaper Fast Mode

    Anthropic has released Claude Opus 4.8, the newest member of its flagship Opus class, available today across every surface and priced exactly like the model it replaces. The company calls it “a modest but tangible improvement” on Opus 4.7, but the framing undersells what is actually interesting here: the headline upgrade is not a benchmark number, it is honesty. Opus 4.8 is built to know when it does not know, and that single behavioral shift may matter more for real agent work than any raw capability bump.

    TLDR

    Claude Opus 4.8 is an across-the-board upgrade to Anthropic’s Opus class that ships today at the same regular price as Opus 4.7 ($5 per million input tokens, $25 per million output tokens), with the model positioned as “a more effective collaborator.” The marquee improvement is honesty: Opus 4.8 is roughly four times less likely than its predecessor to let flaws in its own code pass unremarked, and it is more willing to flag uncertainty rather than confidently claim progress on thin evidence. A pre-release alignment assessment found new highs on prosocial traits like supporting user autonomy and acting in the user’s best interest, with misaligned behavior at rates similar to Anthropic’s best-aligned model, Claude Mythos Preview. Three things launch alongside the model: dynamic workflows in Claude Code (research preview), where Claude plans work then runs hundreds of parallel subagents that run even longer and verify their own outputs before reporting back; effort control in claude.ai and Cowork, a slider for how hard Claude thinks; and a Messages API update that accepts system entries inside the messages array so developers can update instructions mid-task without breaking the prompt cache. Fast mode now runs at 2.5x speed and is three times cheaper than before ($10 / $50 per million tokens). The roadmap points to cheaper Opus-equivalent models, a higher-intelligence class above Opus, and a wider rollout of Mythos-class models gated behind stronger cyber safeguards under Project Glasswing.

    Thoughts

    The most important sentence in this announcement is not about coding scores. It is the claim that Opus 4.8 is about four times less likely than Opus 4.7 to let flaws in its own code slip by without comment. For a chat assistant, overconfidence is annoying. For an agent, it is catastrophic. The whole premise of long-running autonomous work is that you hand the model a task and walk away, which means the model’s own judgment about whether it succeeded becomes the only judgment in the loop until you come back. A model that confidently declares victory on a half-finished migration does not save you time, it costs you a debugging session plus the time you spent trusting it. Honesty, framed this way, is not a soft virtue. It is the load-bearing reliability property that makes unattended agents usable at all.

    Read the launch as a single coherent argument rather than a list of features, and the pieces lock together. Dynamic workflows let Claude plan a job and fan out hundreds of parallel subagents that, with Opus 4.8, run longer than before. Effort control lets you dial up how much the model thinks. The honesty improvement means the model checks its own work and flags what it is unsure about instead of papering over it. Put those three together and you get one product thesis: let it run longer, let it think harder, and trust it to tell you when something is wrong. The codebase-scale migration example, hundreds of thousands of lines from kickoff to merge with the existing test suite as the bar, is the proof point. None of those three capabilities is worth much alone. A model that runs for hours but lies about its results is a liability. A model that flags uncertainty but cannot sustain a long task never reaches the moment where its honesty matters. Anthropic shipped all three at once because they only pay off together.

    The economics deserve a closer look than the “same price” headline invites. Regular pricing is flat versus Opus 4.7, which is the polite way of saying you get a better model for free. The real move is fast mode: 2.5x the speed at three times cheaper than it cost on previous models, landing at $10 per million input and $50 per million output. That is Anthropic quietly attacking the latency-versus-cost tradeoff that has shaped how teams deploy frontier models. Until now, “fast” meant “expensive,” so you reserved it for interactive moments and ate the wait everywhere else. Collapsing that premium changes the default. And note the subtle token story underneath: Opus 4.8 at its default high effort spends roughly the same tokens on coding as Opus 4.7’s default while performing better, so the effort slider is not a way to bleed you dry, it is an honest exposure of the quality-cost dial that was always there implicitly.

    The Messages API change is the kind of unglamorous plumbing that practitioners will appreciate immediately. Letting system entries live inside the messages array means you can update an agent’s instructions, permissions, token budget, or environment context partway through a task without smuggling the update through a fake user turn and without blowing up your prompt cache. Anyone who has built a long-running agent has hit this wall: the world changes mid-task, the agent needs new constraints, and the only clean way to inject them previously was a cache-busting hack. This is Anthropic treating agents as first-class, stateful, long-lived processes rather than oversized chat sessions. It is a small spec change with outsized implications for how you architect an agent that runs for an hour.

    Then there is the roadmap, where the most telling line is the quietest. Anthropic says a small number of organizations are already using Claude Mythos Preview for cybersecurity work under Project Glasswing, and that models of this capability level require stronger cyber safeguards before general release. Notice that they are pinning Opus 4.8’s alignment numbers to Mythos as the benchmark for “best-aligned,” while simultaneously holding Mythos back from general availability on safety grounds. That is a deliberate signal: the next class of model is good enough that they are gating it on cyber-offense risk, not on capability. For a site about the pursuit of joy, fulfillment, and purpose through AI, this is the part worth sitting with. The frontier is increasingly defined not by what the models can do, but by what their builders decide it is responsible to ship. Honesty in the small (flagging a bad line of code) and restraint in the large (holding back a cyber-capable model) are the same instinct expressed at two different scales.

    Key Takeaways

    • Claude Opus 4.8 is now available everywhere, replacing Opus 4.7 as Anthropic’s flagship Opus-class model and positioned as “a more effective collaborator.”
    • Regular usage pricing is unchanged from Opus 4.7, holding at $5 per million input tokens and $25 per million output tokens, so the capability gains come at no added cost.
    • The single most emphasized improvement is honesty, which Anthropic treats as a core trained behavior rather than a marketing flourish.
    • Evaluations show Opus 4.8 is around four times less likely than its predecessor to let flaws in its own code pass unremarked, a direct reliability win for autonomous coding.
    • Early testers report the model is more likely to flag uncertainty about its work and less likely to make unsupported claims or jump to conclusions on thin evidence.
    • A detailed alignment assessment was run before release and concluded Opus 4.8 reaches new highs on prosocial traits like supporting user autonomy and acting in the user’s best interest.
    • Misaligned behavior such as deception or cooperation with misuse is at rates substantially lower than Opus 4.7 and similar to Anthropic’s best-aligned model, Claude Mythos Preview.
    • The full alignment assessment and pre-deployment safety tests are documented in the public Claude Opus 4.8 System Card.
    • Dynamic workflows launch as a research preview inside Claude Code, letting Claude plan the work and then run hundreds of parallel subagents in a single session.
    • With Opus 4.8, those subagents can run even longer, and Claude verifies its outputs before reporting back rather than declaring success blindly.
    • Anthropic’s flagship example for dynamic workflows is a codebase-scale migration across hundreds of thousands of lines of code, from kickoff to merge, using the existing test suite as the success bar.
    • Dynamic workflows are available in Claude Code for the Enterprise, Team, and Max plans.
    • Effort control arrives in claude.ai and Cowork as a setting next to the model selector that lets users choose how much effort Claude puts into a response.
    • Higher effort makes Claude think more frequently and deeply for better answers; lower effort responds faster and consumes rate limits more slowly. Effort control is available on all plans.
    • Opus 4.8 defaults to “high” effort, judged the best overall balance of quality and user experience.
    • On coding tasks, the default effort spends a similar number of tokens as Opus 4.7’s default but delivers better performance, so quality rises without a token penalty.
    • Users can select “extra” (called “xhigh” in Claude Code) or “max” to spend more tokens for stronger results, and Anthropic recommends “extra” for difficult tasks and long-running asynchronous workflows.
    • Rate limits in Claude Code were increased to accommodate the higher token usage of the higher effort levels.
    • The Messages API now accepts system entries inside the messages array, a meaningful change for agent developers.
    • That update lets developers change Claude’s instructions mid-task, adjusting permissions, token budgets, or environment context, without breaking the prompt cache or routing through a user turn.
    • Fast mode now runs at 2.5x speed and is three times cheaper than it was for previous models, priced at $10 per million input tokens and $50 per million output tokens.
    • Developers access the model as claude-opus-4-8 through the Claude API.
    • Partner Miguel Gonzalez reports Opus 4.8 scored 84% on Online-Mind2Web, a meaningful jump over both Opus 4.7 and GPT-5.5, calling it the strongest computer-use and browser-agent model his team has tested.
    • Databricks reports that, inside Genie, Opus 4.8 reasons over unstructured content like PDFs and diagrams at 61% cheaper token cost than Opus 4.7.
    • Thomson Reuters reports Opus 4.8 is the first model to break 10% overall on the all-pass standard of its Legal Agent Benchmark, the highest score recorded there.
    • Eleven partners weighed in, including Cursor, Cognition’s Devin, Databricks Genie, Thomson Reuters CoCounsel, and Hebbia, spanning coding, legal, finance, and enterprise data work.
    • Anthropic is working on models that deliver many of the same capabilities as Opus at a lower cost.
    • The company plans to release a new class of model with even higher intelligence than Opus.
    • Under Project Glasswing, a small number of organizations are already using Claude Mythos Preview for cybersecurity work, with Mythos-class models expected to reach all customers in the coming weeks once stronger cyber safeguards are in place.

    Detailed Summary

    What Claude Opus 4.8 Is

    Claude Opus 4.8 is an upgrade to Anthropic’s Opus class of models, building on Opus 4.7 with improvements across benchmarks covering coding, agentic skills, reasoning, and practical knowledge-work tasks. Anthropic describes the result as “a more effective collaborator” while characterizing the release overall as “a modest but tangible improvement on its predecessor.” The model is available today, everywhere, and developers call it as claude-opus-4-8 via the Claude API. The announcement includes a comparison table against the predecessor and other models, though the per-cell numbers in that table are published as an image and are not reproduced here as text.

    Honesty: The Headline Improvement

    Anthropic singles out honesty as one of the most prominent improvements in Opus 4.8. All of the company’s models are trained to be honest, which includes avoiding claims they cannot support. A persistent problem with AI models generally is that they sometimes jump to conclusions, confidently claiming progress despite thin evidence. Early testers report that Opus 4.8 is more likely to flag uncertainties about its own work and less likely to make unsupported claims. The most concrete measure: evaluations show Opus 4.8 is around four times less likely than its predecessor to allow flaws in code it has written to pass unremarked. For agentic and unattended use, this self-skepticism is the difference between a model that reliably tells you when something went wrong and one that quietly ships a broken result.

    Alignment Assessment

    A detailed alignment assessment was run before release. On the positive side, the Alignment team concluded that Opus 4.8 “reaches new highs on our measures of prosocial traits like supporting user autonomy and acting in the user’s best interest.” On the risk side, misaligned behavior such as deception or cooperation with misuse occurs at rates substantially lower than Opus 4.7, and similar to Anthropic’s best-aligned model, Claude Mythos Preview. The full alignment assessment and the pre-deployment safety tests are published in the Claude Opus 4.8 System Card, which also contains the complete benchmark table and wider evaluations.

    Dynamic Workflows in Claude Code

    Launching today as a research preview in Claude Code, dynamic workflows let Claude plan the work and then run hundreds of parallel subagents in a single session. With Opus 4.8, those agents can run even longer than before, and Claude verifies its outputs before reporting back rather than reporting unchecked results. The showcase example is a codebase-scale migration: Claude Code with Opus 4.8 can carry out migrations across hundreds of thousands of lines of code, all the way from kickoff to merge, using the existing test suite as its bar for success. Dynamic workflows are available in Claude Code for the Enterprise, Team, and Max plans.

    Effort Control

    Effort control arrives in claude.ai and Cowork as a setting alongside the model selector that lets users choose how much effort Claude puts into a response. Higher effort means Claude thinks more frequently and deeply for better responses; lower effort means it responds faster and uses rate limits more slowly. Opus 4.8 defaults to “high” effort, which Anthropic judged the best overall balance of quality and user experience. On coding tasks, that default spends a similar number of tokens as Opus 4.7’s default while performing better. Users who want more can choose “extra” (called “xhigh” in Claude Code) or “max” to spend more tokens for stronger results, and Anthropic recommends “extra” for difficult tasks and long-running asynchronous workflows. To support the heavier token usage at higher effort levels, rate limits in Claude Code were increased. Effort control is available on all plans.

    Messages API Update

    The Messages API now accepts system entries inside the messages array. This lets developers update Claude’s instructions mid-task without breaking the prompt cache and without routing the update through a user turn. In practice that means you can update permissions, token budgets, or environment context while an agent is running, which is exactly the kind of statefulness a long-running autonomous process needs. It is a small specification change with significant consequences for how developers build durable agents.

    Pricing and Fast Mode

    Regular usage pricing is unchanged from Opus 4.7: $5 per million input tokens and $25 per million output tokens. The notable shift is in fast mode, where the model works at 2.5x the speed and fast mode is now three times cheaper than it was for previous models, landing at $10 per million input tokens and $50 per million output tokens. The combination of unchanged regular pricing and dramatically cheaper fast mode reshapes the latency-versus-cost calculus that has long governed how teams deploy frontier models.

    Partner Results Across Coding, Legal, Finance, and Data

    Eleven partners shared results spanning the spectrum of professional work. Miguel Gonzalez reports 84% on Online-Mind2Web, a meaningful jump over both Opus 4.7 and GPT-5.5, calling it the strongest computer-use and browser-agent model his team has tested. Databricks reports that Genie reasons over unstructured content like PDFs and diagrams at 61% cheaper token cost than Opus 4.7. Thomson Reuters reports Opus 4.8 is the first model to break 10% overall on the all-pass standard of its Legal Agent Benchmark. Cursor reports gains across every effort level on CursorBench with more efficient tool calling, and Cognition reports that Devin sees cleaner tool use, fixes to the comment-verbosity and tool-calling issues seen with Opus 4.7, and improvements over Opus 4.6. Hebbia reports strong quality with better citation precision and more token efficiency on retrieval for dense financial filings. The footnotes note that Terminal-Bench 2.1 was scored on the Terminus-2 public harness (GPT-5.5’s Codex CLI harness score is 83.4%), that OSWorld-Verified methodology changed with Opus 4.7’s score updated to 82.3%, and that on Finance Agent v2 Gemini 3.5 Flash scores 57.9%.

    What Is Next: Cheaper Models, Higher Intelligence, and Mythos

    Anthropic outlined a three-part roadmap. First, the company is working on models that provide many of the same capabilities as Opus at a lower cost. Second, it plans to release a new class of model with even higher intelligence than Opus. Third, as part of Project Glasswing, a small number of organizations are currently using Claude Mythos Preview for cybersecurity work; models of this capability level require stronger cyber safeguards before general release, and Anthropic expects to bring Mythos-class models to all customers in the coming weeks.

    Notable Quotes

    “Claude Opus 4.8 has noticeably better judgment. In Claude Code, it asks the right questions, catches its own mistakes, pushes back when a plan isn’t sound, and builds up confidence around complex, multi-service explorations before making big changes. It’s a great model to build with.”

    Tom Pritchard, Staff Engineer, in Claude Code

    “On our Super-Agent benchmark, Claude Opus 4.8 is the only model to complete every case end-to-end, beating prior Opus models and GPT-5.5 at parity on cost. For agent products in translation, deep research, slide-building, and analysis, it delivers powerful reliability.”

    Kay Zhu, Co-Founder and CTO, on the Super-Agent benchmark

    “On CursorBench, Claude Opus 4.8 exceeds prior Opus models across every effort level. Tool calling is meaningfully more efficient, using fewer steps for the same intelligence, and it carries end-to-end tasks through.”

    Michael Truell, Co-Founder and CEO, on CursorBench results

    “Claude Opus 4.8 delivers the highest score recorded on our Legal Agent Benchmark, and is the first model to break 10% overall on the all-pass standard. For substantive legal work, that’s the kind of accuracy lift that translates directly into how much real attorney work our customers can hand off with confidence.”

    Niko Grupen, Head of Applied Research, on the Legal Agent Benchmark

    “Claude Opus 4.8 feels like a major quality-of-life update over Opus 4.7: faster, easier to collaborate with, and better at carrying context and style direction across a long session. Opus 4.8 is the model I kept trusting for work where voice, taste, and technical execution all have to happen side-by-side.”

    Katie Parrott, Staff Writer, on long writing sessions

    “Claude Opus 4.8 is the strongest computer-use and browser-agent model we’ve tested, scoring 84% on Online-Mind2Web, which is a meaningful jump over both Opus 4.7 and GPT-5.5. It stays reflective and on-task in the way our customers’ agent workloads need to be reliable end-to-end.”

    Miguel Gonzalez, Tech Lead, on computer-use and browser agents

    “Claude Opus 4.8 uses tools cleanly and follows instructions with the consistency our autonomous engineering workloads need to keep running unattended. It improves on Opus 4.6 and fixes the comment-verbosity and tool-calling issues we saw with Opus 4.7. This release from Anthropic translates directly into faster capability gains for engineers building on Devin.”

    Scott Wu, CEO, on building with Devin

    “On our long-running evals, Claude Opus 4.8’s analysis was consistently higher quality than prior Opus models. It finished faster and produced richer, more information dense outputs. Overall, a noticeably better signal to noise ratio. The biggest differentiator was Opus 4.8’s tendency to proactively flag issues with the inputs and outputs of an analysis, something other models routinely missed and left to the users to catch.”

    Michael Ran, Sr. Investment Associate, on long-running analysis evals

    Claude Opus 4.8 is a quieter release than its “modest but tangible” billing suggests, because the gains land where autonomous work actually lives: a model that flags its own uncertainty, runs longer and checks itself, scales effort on demand, and stays affordable while fast mode gets cheaper. The honesty improvement alone changes the trust math for anyone deploying agents. Read Anthropic’s full announcement here.

    Related Reading

  • Claude Opus 4.7 Released: Anthropic’s New Coding Powerhouse With xhigh Effort Mode, 3.75MP Vision, and State-of-the-Art Agentic Performance

    TLDR

    Anthropic released Claude Opus 4.7 on April 16, 2026, as a direct upgrade to Opus 4.6. It delivers major gains on the hardest coding tasks, introduces a new xhigh effort level, supports images up to 2,576 pixels on the long edge (roughly 3.75 megapixels), and ships with automatic cybersecurity safeguards. Pricing stays flat at $5 per million input tokens and $25 per million output tokens. Early testers at Cursor, Replit, Vercel, Notion, Devin, Harvey, Databricks, and Warp report double-digit benchmark jumps, stronger instruction following, better long-horizon autonomy, and a more opinionated model that pushes back instead of agreeing reflexively.

    Key Takeaways

    • Direct upgrade from Opus 4.6 at the same price point, available via API as claude-opus-4-7, plus Amazon Bedrock, Google Cloud Vertex AI, and Microsoft Foundry.
    • New xhigh effort level slots between high and max, giving developers finer control over the reasoning-versus-latency tradeoff.
    • Vision gets a real jump: images up to 2,576 pixels on the long edge, more than 3x prior Claude models. XBOW reported 98.5% visual acuity versus 54.5% for Opus 4.6.
    • Coding benchmarks up across the board: Cursor saw 70% on CursorBench versus 58% for 4.6, Rakuten-SWE-Bench resolved 3x more production tasks, and GitHub measured a 13% lift on their 93-task benchmark.
    • Long-horizon autonomy is a headline theme. Devin says Opus 4.7 works coherently for hours. Genspark highlights loop resistance and the highest quality-per-tool-call ratio they have measured.
    • Instruction following is substantially tighter, which means old prompts written for loose-interpretation models may now behave unexpectedly. Re-tune prompts and harnesses.
    • Better memory across file-system-based workflows, reducing the need for up-front context in multi-session work.
    • Tokenizer changed: same input can now map to 1.0 to 1.35x more tokens. Opus 4.7 also thinks more at higher effort levels, so output token counts rise too.
    • Cybersecurity safeguards automatically detect and block prohibited or high-risk cyber requests. Legitimate security researchers can apply to the new Cyber Verification Program.
    • Claude Code gets /ultrareview, a dedicated review session that catches bugs and design issues. Pro and Max users get three free ultrareviews. Auto mode is extended to Max users.
    • State-of-the-art on GDPval-AA, a third-party evaluation of economically valuable knowledge work spanning finance, legal, and other domains.
    • Not the most capable overall model. That distinction still goes to Claude Mythos Preview, which also remains the best-aligned model Anthropic has trained.

    Detailed Summary

    What Claude Opus 4.7 Actually Is

    Claude Opus 4.7 is Anthropic’s latest generally available frontier model, positioned as a targeted upgrade to Opus 4.6 rather than a ground-up new generation. The focus is squarely on advanced software engineering, long-running agentic workflows, and higher-fidelity vision. Anthropic describes it as handling complex, long-running tasks with rigor and consistency, paying precise attention to instructions, and devising ways to verify its own outputs before reporting back.

    The positioning matters. Claude Mythos Preview, announced alongside Project Glasswing, remains the most powerful and best-aligned model Anthropic has trained. Opus 4.7 is the first release after Mythos Preview and serves a dual purpose: give developers a concrete upgrade today, and stress-test new cybersecurity safeguards on a less capable model before Anthropic attempts a broader release of Mythos-class systems.

    Coding and Agentic Performance

    The early-access testimonials read like a highlight reel of the agentic coding ecosystem. Cursor saw CursorBench scores jump from 58% on Opus 4.6 to over 70% on Opus 4.7. Rakuten measured 3x more resolved production tasks on Rakuten-SWE-Bench with double-digit gains in code quality and test quality. GitHub measured a 13% lift on a 93-task coding benchmark including four tasks that neither Opus 4.6 nor Sonnet 4.6 could solve. Notion observed a 14% improvement over Opus 4.6 at fewer tokens and a third of the tool errors, calling it the first model to pass their implicit-need tests.

    Devin emphasized sustained autonomy, saying the model works coherently for hours and pushes through hard problems rather than giving up. Warp reported that Opus 4.7 passed Terminal Bench tasks prior Claude models had failed, including a tricky concurrency bug Opus 4.6 could not crack. Vercel highlighted a behavior they had not seen before: the model actually does proofs on systems code before starting work, and is noticeably more honest about its own limits.

    A recurring theme across testimonials is that Opus 4.7 pushes back. Replit’s president said it feels like a better coworker because it challenges technical decisions instead of agreeing by default. Augment Code noted it brings a more opinionated perspective rather than simply agreeing with the user. For anyone building real engineering workflows, that pushback behavior is arguably more valuable than raw benchmark deltas.

    Vision: The Quiet Breakthrough

    The vision upgrade may be the most underappreciated change. Opus 4.7 now accepts images up to 2,576 pixels on the long edge, roughly 3.75 megapixels, which is more than three times the previous Claude limit. This is a model-level change, not an API parameter, so every image sent to Claude is processed at higher fidelity automatically.

    XBOW, which builds autonomous penetration testing agents that rely heavily on computer use, reported the most dramatic single number in the entire announcement: 98.5% on their visual acuity benchmark versus 54.5% for Opus 4.6. They described their single biggest Opus pain point as effectively disappearing, unlocking an entire class of work where they could not previously use Claude. Solve Intelligence reported major improvements in multimodal understanding for life sciences patent workflows, from reading chemical structures to interpreting complex technical diagrams.

    This unlocks computer-use agents reading dense screenshots, data extraction from complex diagrams, and any work requiring pixel-perfect references.

    The New xhigh Effort Level

    Opus 4.7 introduces an xhigh (extra high) effort level that sits between high and max. This gives developers a new middle gear for the reasoning-versus-latency tradeoff on hard problems. In Claude Code, Anthropic raised the default effort level to xhigh across all plans. For coding and agentic use cases, Anthropic recommends starting with high or xhigh effort rather than defaulting to medium.

    Alongside effort controls, the Claude Platform is getting task budgets in public beta, letting developers guide Claude’s token spend so it can prioritize work across longer runs. This matters because Opus 4.7 thinks more at higher effort levels, particularly on later turns in agentic settings.

    Token Usage Changes You Need to Plan For

    Two token-related changes affect migration. First, Opus 4.7 uses an updated tokenizer that improves how the model processes text, but the tradeoff is that the same input can map to 1.0 to 1.35x more tokens depending on content type. Second, Opus 4.7 thinks more at higher effort levels, which means more output tokens on hard problems.

    Anthropic’s own internal coding evaluation shows the net effect is favorable when measured against quality delivered per token, but the recommendation is to measure the difference on real traffic rather than assume. Token usage can be controlled via the effort parameter, task budgets, or simply prompting the model to be more concise. Anthropic published a migration guide with tuning advice.

    Claude Code Updates: /ultrareview and Auto Mode

    Claude Code gets two meaningful additions. The new /ultrareview slash command produces a dedicated review session that reads through changes and flags bugs and design issues that a careful reviewer would catch. Pro and Max users get three free ultrareviews to try it out.

    Auto mode, a permissions option where Claude makes decisions on behalf of the user so longer tasks run with fewer interruptions, has been extended from Pro to Max users. The pitch is that auto mode is safer than skipping all permissions while still enabling long autonomous runs.

    Cybersecurity Safeguards and the Cyber Verification Program

    Opus 4.7 ships with safeguards that automatically detect and block requests indicating prohibited or high-risk cybersecurity uses. During training, Anthropic experimented with efforts to differentially reduce cyber capabilities, meaning Opus 4.7’s cyber ceiling is intentionally lower than Mythos Preview’s.

    For legitimate users, Anthropic launched a Cyber Verification Program for security professionals doing vulnerability research, penetration testing, and red-teaming. Real-world data from these safeguards will inform how Anthropic eventually releases Mythos-class models more broadly.

    Safety and Alignment

    Opus 4.7 shows a similar safety profile to Opus 4.6 overall. Honesty and resistance to prompt injection attacks improved. Some measures slipped modestly, notably a tendency to give overly detailed harm-reduction advice on controlled substances. Anthropic’s alignment assessment concluded the model is largely well-aligned and trustworthy, though not fully ideal. Mythos Preview still holds the crown as the best-aligned model according to Anthropic’s evaluations. The full Claude Opus 4.7 System Card has the complete breakdown.

    Real-World Work Beyond Code

    Opus 4.7 posts a state-of-the-art score on the Finance Agent evaluation and on GDPval-AA, a third-party evaluation of economically valuable knowledge work spanning finance, legal, and other domains. Harvey reported 90.9% on BigLaw Bench at high effort with noticeably smarter handling of ambiguous document editing tasks, including correctly distinguishing assignment provisions from change-of-control provisions. Databricks measured 21% fewer errors than Opus 4.6 on OfficeQA Pro document reasoning. Vercel went as far as calling it the best model in the world for building dashboards and data-rich interfaces.

    Pricing and Availability

    Pricing holds at $5 per million input tokens and $25 per million output tokens. Opus 4.7 is live today across all Claude products, the Claude API as claude-opus-4-7, Amazon Bedrock, Google Cloud Vertex AI, and Microsoft Foundry.

    Thoughts

    The most interesting thing about this release is not the benchmark deltas, which are strong but expected for a point-release. It is the behavioral shift. When a dozen independent companies describe the same model as opinionated, willing to push back, self-verifying, and honest about its limits, that is a different product category than “next version, slightly better.” That is a model optimized for being a collaborator rather than an autocomplete.

    For solo builders running long agentic sessions, the loop resistance and long-horizon autonomy claims are the ones worth taking seriously. Genspark’s framing is sharp: a model that loops indefinitely on 1 in 18 queries wastes compute and blocks users. If Opus 4.7 genuinely closes that failure mode, the economics of overnight autonomous runs change meaningfully.

    The vision jump is the sleeper feature. 3.75 megapixel support plus the XBOW acuity number suggests computer-use agents are about to get a lot more reliable at reading actual screens. Anyone building browser agents, automated QA, or visual data extraction pipelines should retest their stacks this week.

    The instruction-following tightening is a real gotcha. Prompts written against Opus 4.6’s looser interpretation habits may produce surprising results when the model now takes every word literally. Teams with production prompt libraries should budget time for re-tuning rather than expecting a drop-in swap.

    Finally, the strategic framing around Mythos Preview is worth noting. Anthropic is explicitly using Opus 4.7 as a safeguards testbed for eventually releasing more capable cyber-capable systems. That is an honest acknowledgment that capability and deployment readiness are separate problems, and it sets a template for how frontier releases may work going forward.