none reviewed yet · 2 benchmark observations · latest capture Sep 8, 2026
Inspect the evidenceAnthropic · MODEL PROFILE
Claude Opus 5
Production teams that want near-flagship correctness at half of Fable pricing
Saved editorial fit for your current task lens; not a benchmark result
How scoring worksVerified rates (Sep 8, 2026) · 100 runs / month at your token settings
Adjust workloadAt a glance
9 facts- Provider
- Anthropic
- Type
- API model
- Access
- Anthropic API, Claude apps, AWS Bedrock, Claude Code
- Context
- 1,000,000 tokens
- Maximum output
- 128K tokens
- License / weights
- Proprietary
- Input modalities
- text, image, file
- Output modalities
- text
- Pricing notes
- $5 in / $25 out per 1M; cache read $0.50, cache write $6.25
The practical trade-offs
Documented strengths
- SWE-bench Verified 96% is the single highest coding score in this evidence base
- Best AA Coding Agent Index v1.4 (68.1) of any model in OpenAI's own table
- Half the token price of Fable 5.1 with a large, well-sampled Arena row (8,431 votes)
- Documented frontend QA behaviour: opened desktop and phone widths, found and fixed an off-screen checkout button
Limitations to check
- $5/$25 is still premium pricing
- TTFT 77.82 s at max effort
- Fable 5.1 beats it on SWE-bench Pro, Terminal-Bench 4.0, OSWorld and Design Arena
- Effort tier changes its Arena rank by 28 Elo (max vs high)
Follow the source
7 source linksEvidence & conditions
7Source links
NoneReviewed records
2Benchmark runs
Sep 8, 2026Latest capture
- Evidence source 1
anthropic.comDate not recorded - Evidence source 2
docs.aws.amazon.comDate not recorded - Evidence source 3
anthropic.comDate not recorded - Evidence source 4
arena.aiDate not recorded - Evidence source 5
benchlm.aiDate not recorded - Evidence source 6
benchlm.aiDate not recorded - Evidence source 7
artificialanalysis.aiDate not recorded
Benchmark observations
Terminal-Bench 4.0 Inspect conditions
- name
- Terminal-Bench 4.0
- value
- 52.6
- unit
- %
- source Url
- https://openai.com/index/gpt-6-astra/
- checked At
- 2026-09-08T19:30:56Z
- attribution
- provider-reported
- configuration
- OpenAI comparative table, not an independent Anthropic verification; maximum across effort; competitor harness not specified.
DeepSWE v1.1 Inspect conditions
- name
- DeepSWE v1.1
- value
- 73.7
- unit
- %
- source Url
- https://openai.com/index/gpt-6-astra/
- checked At
- 2026-09-08T19:30:56Z
- attribution
- provider-reported
- configuration
- OpenAI comparative table, not an independent Anthropic verification; maximum across effort; competitor harness not specified.
Source reports, measured results, and editorial assessments are separate evidence types. Source dates do not imply a new benchmark run. Capture dates are shown when recorded; otherwise we say the capture date was not recorded. Recorded dates are compared with this catalogue snapshot (Sep 11, 2026), not today’s date.
Method AIS-V2
How this page is assessed
Quality claims follow Method AIS-V2.
Money never buys rank — commercial relationships cannot change rank, eligibility, evidence quality, or publication timing.
Submit or claim a listing is free and does not skip the editorial queue.
Private · this device only