Claude Haiku 5.5
- Input price
- $0.10
- Per 1 million tokens (USD)
- Output price
- $0.50
- Per 1 million tokens (USD)
- Effort levels
- 5total
- Default: Med
- Benchmark results
- 40total
- Benchmark types: 22
Model information
- API model ID
- claude-haiku-5-5
- release date
- 2026-10-07
- effort
- Low / Med (default) / High / Xhigh / Max
- Effort source
- Based on Anthropic effort documentation (platform.claude.com/docs/en/build-with-claude/effort).
- Claude Code
- available
- Codex
- unavailable
- Availability notes
- `haiku` resolves to Haiku 5.5 on Anthropic API (as of v2.1.293). Claude Code default effort is medium. Codex's official model list (learn.chatgpt.com/docs/models) contains only OpenAI models, so codex is false (connections via a custom provider are not counted as official options).
- Availability source
- https://code.claude.com/docs/en/model-config
Price (USD per 1 million tokens)
Prices do not vary by effort; token consumption does.
| input | $0.10 |
|---|---|
| output | $0.50 |
| cache read | $0.01 |
| cache write | $0.13 |
cacheWritePerMTok is for a 5-minute cache write. Values are for prompts of 100,000 tokens or fewer. Above 100,000 tokens: input $0.50 / 5m write $0.625 / 1h write $1 / cache read $0.05 / output $2.50.
Claude Haiku 5.5 (for prompts up to 100,000 tokens) | $0.10 / MTok | $0.125 / MTok | $0.20 / MTok | $0.01 / MTok | $0.50 / MTokSource: https://platform.claude.com/docs/en/about-claude/pricing
Benchmark results
| Benchmark | effort | Score | Source cost (metric) | measured/reported by | source | Conditions and extraction method |
|---|---|---|---|---|---|---|
| Terminal-Bench 4.0Low | Low | 12.7% | $0.42Cost per attempt | Anthropic | Introducing Claude Haiku 5.5official · Retrieved 2026-11-12 | Vendor-reported value. A point in the figure "Terminal-Bench 4.0 / Accuracy vs. cost" on the Anthropic announcement page. values were taken from each point's aria-label (DOM); the page shows them only in the figure; value for its own model (announced by Anthropic)transcribed from textSource excerptTerminal-Bench 4.0 — Accuracy vs. cost: "Haiku 5.5 · Low: 12.7% $0.42" |
| Terminal-Bench 4.0Med (default) | Med | 20.3% | $0.68Cost per attempt | Anthropic | Introducing Claude Haiku 5.5official · Retrieved 2026-11-12 | Vendor-reported value. A point in the figure "Terminal-Bench 4.0 / Accuracy vs. cost" on the Anthropic announcement page. values were taken from each point's aria-label (DOM); the page shows them only in the figure; value for its own model (announced by Anthropic)transcribed from textSource excerptTerminal-Bench 4.0 — Accuracy vs. cost: "Haiku 5.5 · Med: 20.3% $0.68" |
| Terminal-Bench 4.0High | High | 24.8% | $1.04Cost per attempt | Anthropic | Introducing Claude Haiku 5.5official · Retrieved 2026-11-12 | Vendor-reported value. A point in the figure "Terminal-Bench 4.0 / Accuracy vs. cost" on the Anthropic announcement page. values were taken from each point's aria-label (DOM); the page shows them only in the figure; value for its own model (announced by Anthropic)transcribed from textSource excerptTerminal-Bench 4.0 — Accuracy vs. cost: "Haiku 5.5 · High: 24.8% $1.04" |
| Terminal-Bench 4.0Xhigh | Xhigh | 31.5% | $1.75Cost per attempt | Anthropic | Introducing Claude Haiku 5.5official · Retrieved 2026-11-12 | Vendor-reported value. A point in the figure "Terminal-Bench 4.0 / Accuracy vs. cost" on the Anthropic announcement page. values were taken from each point's aria-label (DOM); the page shows them only in the figure; value for its own model (announced by Anthropic)transcribed from textSource excerptTerminal-Bench 4.0 — Accuracy vs. cost: "Haiku 5.5 · Xhigh: 31.5% $1.75" |
| Terminal-Bench 4.0Max | Max | 39.2% | $2.64Cost per attempt | Anthropic | Introducing Claude Haiku 5.5official · Retrieved 2026-11-12 | Vendor-reported value. A point in the figure "Terminal-Bench 4.0 / Accuracy vs. cost" on the Anthropic announcement page. values were taken from each point's aria-label (DOM); the page shows them only in the figure; value for its own model (announced by Anthropic)transcribed from textSource excerptTerminal-Bench 4.0 — Accuracy vs. cost: "Haiku 5.5 · Max: 39.2% $2.64" |
| Terminal-Bench-Science 0.1Max | Max | 20.6% | — | Anthropic | Claude Haiku 5.5 System Cardofficial · Retrieved 2026-11-14 | Vendor-reported value. Claude Code --bare, max, 70 tasks. No internet; safeguards enabled. No fallback; safeguards-triggered exits count as failures (2.3% of 700 trials). Cost not stated.transcribed from textSource excerpt20.6%Location §8.5 “Terminal-Bench-Science 0.1”, body text, own model score |
| FrontierCode v1.1(main set)not stated | not stated | 46.4% | — | Anthropic | Introducing Claude Haiku 5.5official · Retrieved 2026-11-12 | Vendor-reported value. table note: see the Haiku 5.5 System Card (effort not stated)transcribed from tableSource excerpt46.4%Location Table, FrontierCode v1.1 (Main) row: 46.4 |
| FrontierSWE 2Max | Max | 43.8% | — | Proximal | Claude Haiku 5.5 System Cardofficial · Retrieved 2026-11-14 | Vendor-reported value. Run by Proximal in its own agent harness. max, 34 tasks, averaged over 5 trials per task. Cost not stated.transcribed from textSource excerpt43.8%Location §8.6 “FrontierSWE v2”, body text, own model score |
| GDPval-AA v2.1Low | Low | 1,125 | $0.01Cost per task | Anthropic | Introducing Claude Haiku 5.5official · Retrieved 2026-11-12 | Vendor-reported value. A point in the figure "GDPval-AA v2.1 / Accuracy vs. cost" on the Anthropic announcement page. values were taken from each point's aria-label (DOM); the page shows them only in the figure; value for its own model (announced by Anthropic)transcribed from textSource excerptGDPval-AA v2.1 — Accuracy vs. cost: "Haiku 5.5 · Low: 1125.0 $0.01" |
| GDPval-AA v2.1Med (default) | Med | 1,277 | $0.03Cost per task | Anthropic | Introducing Claude Haiku 5.5official · Retrieved 2026-11-12 | Vendor-reported value. A point in the figure "GDPval-AA v2.1 / Accuracy vs. cost" on the Anthropic announcement page. values were taken from each point's aria-label (DOM); the page shows them only in the figure; value for its own model (announced by Anthropic)transcribed from textSource excerptGDPval-AA v2.1 — Accuracy vs. cost: "Haiku 5.5 · Med: 1277.0 $0.03" |
| GDPval-AA v2.1High | High | 1,420 | $0.09Cost per task | Anthropic | Introducing Claude Haiku 5.5official · Retrieved 2026-11-12 | Vendor-reported value. A point in the figure "GDPval-AA v2.1 / Accuracy vs. cost" on the Anthropic announcement page. values were taken from each point's aria-label (DOM); the page shows them only in the figure; value for its own model (announced by Anthropic)transcribed from textSource excerptGDPval-AA v2.1 — Accuracy vs. cost: "Haiku 5.5 · High: 1420.0 $0.09" |
| GDPval-AA v2.1Xhigh | Xhigh | 1,513 | $0.27Cost per task | Anthropic | Introducing Claude Haiku 5.5official · Retrieved 2026-11-12 | Vendor-reported value. A point in the figure "GDPval-AA v2.1 / Accuracy vs. cost" on the Anthropic announcement page. values were taken from each point's aria-label (DOM); the page shows them only in the figure; value for its own model (announced by Anthropic)transcribed from textSource excerptGDPval-AA v2.1 — Accuracy vs. cost: "Haiku 5.5 · Xhigh: 1513.0 $0.27" |
| GDPval-AA v2.1Max | Max | 1,620 | $0.87Cost per task | Anthropic | Introducing Claude Haiku 5.5official · Retrieved 2026-11-12 | Vendor-reported value. A point in the figure "GDPval-AA v2.1 / Accuracy vs. cost" on the Anthropic announcement page. values were taken from each point's aria-label (DOM); the page shows them only in the figure; value for its own model (announced by Anthropic)transcribed from textSource excerptGDPval-AA v2.1 — Accuracy vs. cost: "Haiku 5.5 · Max: 1620.0 $0.87" |
| AA-Briefcase v1.1(Elo)not stated | not stated | 1,578 | — | Anthropic | Introducing Claude Haiku 5.5official · Retrieved 2026-11-12 | Vendor-reported value. table note: see the Haiku 5.5 System Card (effort not stated)transcribed from tableSource excerpt1578Location Table, AA-Briefcase v1.1 (Elo) row: 1578 |
| Humanity's Last Exam (with tools)not stated | not stated | 57.4% | — | Anthropic | Introducing Claude Haiku 5.5official · Retrieved 2026-11-12 | Vendor-reported value. table note: see the Haiku 5.5 System Card (effort not stated)transcribed from tableSource excerpt57.4%Location Table, Humanity's Last Exam (with tools) row: 57.4 |
| Humanity's Last Exam (without tools)Low | Low | 30.9% | $0.003Cost per attempt | Anthropic | Introducing Claude Haiku 5.5official · Retrieved 2026-11-12 | Vendor-reported value. A point in the figure "Humanity’s Last Exam (no tools) / Accuracy vs. cost" on the Anthropic announcement page. values were taken from each point's aria-label (DOM); the page shows them only in the figure; value for its own model (announced by Anthropic)transcribed from textSource excerptHumanity’s Last Exam (no tools) — Accuracy vs. cost: "Haiku 5.5 · Low: 30.9% $0.003" |
| Humanity's Last Exam (without tools)Med (default) | Med | 35.5% | $0.007Cost per attempt | Anthropic | Introducing Claude Haiku 5.5official · Retrieved 2026-11-12 | Vendor-reported value. A point in the figure "Humanity’s Last Exam (no tools) / Accuracy vs. cost" on the Anthropic announcement page. values were taken from each point's aria-label (DOM); the page shows them only in the figure; value for its own model (announced by Anthropic)transcribed from textSource excerptHumanity’s Last Exam (no tools) — Accuracy vs. cost: "Haiku 5.5 · Med: 35.5% $0.007" |
| Humanity's Last Exam (without tools)High | High | 39.7% | $0.02Cost per attempt | Anthropic | Introducing Claude Haiku 5.5official · Retrieved 2026-11-12 | Vendor-reported value. A point in the figure "Humanity’s Last Exam (no tools) / Accuracy vs. cost" on the Anthropic announcement page. values were taken from each point's aria-label (DOM); the page shows them only in the figure; value for its own model (announced by Anthropic)transcribed from textSource excerptHumanity’s Last Exam (no tools) — Accuracy vs. cost: "Haiku 5.5 · High: 39.7% $0.02" |
| Humanity's Last Exam (without tools)Xhigh | Xhigh | 44.1% | $0.06Cost per attempt | Anthropic | Introducing Claude Haiku 5.5official · Retrieved 2026-11-12 | Vendor-reported value. A point in the figure "Humanity’s Last Exam (no tools) / Accuracy vs. cost" on the Anthropic announcement page. values were taken from each point's aria-label (DOM); the page shows them only in the figure; value for its own model (announced by Anthropic)transcribed from textSource excerptHumanity’s Last Exam (no tools) — Accuracy vs. cost: "Haiku 5.5 · Xhigh: 44.1% $0.06" |
| Humanity's Last Exam (without tools)Max | Max | 45.9% | $0.21Cost per attempt | Anthropic | Introducing Claude Haiku 5.5official · Retrieved 2026-11-12 | Vendor-reported value. A point in the figure "Humanity’s Last Exam (no tools) / Accuracy vs. cost" on the Anthropic announcement page. values were taken from each point's aria-label (DOM); the page shows them only in the figure; value for its own model (announced by Anthropic)transcribed from textSource excerptHumanity’s Last Exam (no tools) — Accuracy vs. cost: "Haiku 5.5 · Max: 45.9% $0.21" |
| OSWorld 2.1(offline subset)Low | Low | 42.0% | $0.07Cost per attempt | Anthropic | Introducing Claude Haiku 5.5official · Retrieved 2026-11-12 | Vendor-reported value. A point in the figure "OSWorld 2.1 (offline subset) / Accuracy vs. cost" on the Anthropic announcement page. values were taken from each point's aria-label (DOM); the page shows them only in the figure; value for its own model (announced by Anthropic)transcribed from textSource excerptOSWorld 2.1 (offline subset) — Accuracy vs. cost: "Haiku 5.5 · Low: 42.0% $0.07" |
| OSWorld 2.1(offline subset)Med (default) | Med | 53.3% | $0.13Cost per attempt | Anthropic | Introducing Claude Haiku 5.5official · Retrieved 2026-11-12 | Vendor-reported value. A point in the figure "OSWorld 2.1 (offline subset) / Accuracy vs. cost" on the Anthropic announcement page. values were taken from each point's aria-label (DOM); the page shows them only in the figure; value for its own model (announced by Anthropic)transcribed from textSource excerptOSWorld 2.1 (offline subset) — Accuracy vs. cost: "Haiku 5.5 · Med: 53.3% $0.13" |
| OSWorld 2.1(offline subset)High | High | 61.3% | $0.18Cost per attempt | Anthropic | Introducing Claude Haiku 5.5official · Retrieved 2026-11-12 | Vendor-reported value. A point in the figure "OSWorld 2.1 (offline subset) / Accuracy vs. cost" on the Anthropic announcement page. values were taken from each point's aria-label (DOM); the page shows them only in the figure; value for its own model (announced by Anthropic)transcribed from textSource excerptOSWorld 2.1 (offline subset) — Accuracy vs. cost: "Haiku 5.5 · High: 61.3% $0.18" |
| OSWorld 2.1(offline subset)Xhigh | Xhigh | 67.6% | $0.28Cost per attempt | Anthropic | Introducing Claude Haiku 5.5official · Retrieved 2026-11-12 | Vendor-reported value. A point in the figure "OSWorld 2.1 (offline subset) / Accuracy vs. cost" on the Anthropic announcement page. values were taken from each point's aria-label (DOM); the page shows them only in the figure; value for its own model (announced by Anthropic)transcribed from textSource excerptOSWorld 2.1 (offline subset) — Accuracy vs. cost: "Haiku 5.5 · Xhigh: 67.6% $0.28" |
| OSWorld 2.1(offline subset)Max | Max | 72.4% | $0.61Cost per attempt | Anthropic | Introducing Claude Haiku 5.5official · Retrieved 2026-11-12 | Vendor-reported value. A point in the figure "OSWorld 2.1 (offline subset) / Accuracy vs. cost" on the Anthropic announcement page. values were taken from each point's aria-label (DOM); the page shows them only in the figure; value for its own model (announced by Anthropic)transcribed from textSource excerptOSWorld 2.1 (offline subset) — Accuracy vs. cost: "Haiku 5.5 · Max: 72.4% $0.61" |
| Chartography (with tools)Max | Max | 86.2% | — | Anthropic | Claude Haiku 5.5 System Cardofficial · Retrieved 2026-11-14 | Vendor-reported value. Run internally by Anthropic. adaptive thinking, max, with tools, averaged over 5 runs. The grader is Gemini 3.5 Flash. Cost not stated.transcribed from textSource excerpt86.2%Location §8.9.1 “Chartography”, body text, own model with tools score |
| Chartography (without tools)not stated | not stated | 46.4% | — | Anthropic | Introducing Claude Haiku 5.5official · Retrieved 2026-11-12 | Vendor-reported value. table note: see the Haiku 5.5 System Card (effort not stated)transcribed from tableSource excerpt46.4%Location Table, Chartography (no tools) row: 46.4 |
| HealthBench Professional (length-adjusted score)Max | Max | 58.4 | — | Artificial Analysis | Claude Fable 5.2 and Claude Mythos 5.2 System Cardofficial · Retrieved 2026-11-14 | Vendor-reported value. Independently run by Artificial Analysis at Anthropic’s request. max, length-adjusted score, GPT-5.4 low grader. fallback status is not stated for this model. Not comparable with internal runs in earlier Claude system cards (Opus 4.8 grader). Cost not stated.transcribed from textSource excerpt58.4%Location §8.16.2 “HealthBench Professional (Artificial Analysis)”, body text, the model’s length-adjusted score in parentheses |
| GPQA DiamondMax | Max | 89.7% | — | Epoch AI | Epoch AI, Capabilities & benchmarking (benchmark_data.zip, gpqa_diamond.csv) CC BY 4.0third-party · Retrieved 2026-11-12 | Run by Epoch AI (third-party measurement). Source label "Claude Haiku 5.5 (max)". effort is indicated by the suffix _max in Epoch's Model version. Data retrieved 2026-11-12.transcribed from tableSource excerptgpqa_diamond.csv: Model version=claude-haiku-5-5_max, Best score (across scorers)=0.8964646464646465, stderr=0.019774484382483127, Started at=2026-10-09T14:35:53.000Z |
| FrontierMath Tiers 1-3(v2)Max | Max | 75.1% | — | Epoch AI | Epoch AI, Capabilities & benchmarking (benchmark_data.zip, frontiermath_tiers_1_3_v2.csv) CC BY 4.0third-party · Retrieved 2026-11-12 | Run by Epoch AI (third-party measurement). Source label "Claude Haiku 5.5 (max)". effort is indicated by the suffix _max in Epoch's Model version. Data retrieved 2026-11-12.transcribed from tableSource excerptfrontiermath_tiers_1_3_v2.csv: Model version=claude-haiku-5-5_max, Best score (across scorers)=0.7508771929824561, stderr=0.02566445410232946, Started at=2026-10-07T20:42:31.000Z |
| FrontierMath Tier 4(v2)Max | Max | 46.3% | — | Epoch AI | Epoch AI, Capabilities & benchmarking (benchmark_data.zip, frontiermath_tier_4_v2.csv) CC BY 4.0third-party · Retrieved 2026-11-12 | Run by Epoch AI (third-party measurement). Source label "Claude Haiku 5.5 (max)". effort is indicated by the suffix _max in Epoch's Model version. Data retrieved 2026-11-12.transcribed from tableSource excerptfrontiermath_tier_4_v2.csv: Model version=claude-haiku-5-5_max, Best score (across scorers)=0.4634146341463415, stderr=0.07884502378168105, Started at=2026-10-07T20:42:31.000Z |
| OTIS Mock AIME 2024-2025High | High | 97.2% | — | Epoch AI | Epoch AI, Capabilities & benchmarking (benchmark_data.zip, otis_mock_aime_2024_2025.csv) CC BY 4.0third-party · Retrieved 2026-11-12 | Run by Epoch AI (third-party measurement). Source label "Claude Haiku 5.5 (high)". effort is indicated by the suffix _high in Epoch's Model version. Data retrieved 2026-11-12.transcribed from tableSource excerptotis_mock_aime_2024_2025.csv: Model version=claude-haiku-5-5_high, Best score (across scorers)=0.9722222222222222, stderr=0.013109441770118595, Started at=2026-10-07T21:16:25.000Z |
| OTIS Mock AIME 2024-2025Xhigh | Xhigh | 98.9% | — | Epoch AI | Epoch AI, Capabilities & benchmarking (benchmark_data.zip, otis_mock_aime_2024_2025.csv) CC BY 4.0third-party · Retrieved 2026-11-12 | Run by Epoch AI (third-party measurement). Source label "Claude Haiku 5.5 (xhigh)". effort is indicated by the suffix _xhigh in Epoch's Model version. Data retrieved 2026-11-12.transcribed from tableSource excerptotis_mock_aime_2024_2025.csv: Model version=claude-haiku-5-5_xhigh, Best score (across scorers)=0.9888888888888889, stderr=0.005362818328685681, Started at=2026-10-07T21:16:36.000Z |
| SimpleQA VerifiedMax | Max | 23.8% | — | Epoch AI | Epoch AI, Capabilities & benchmarking (benchmark_data.zip, simpleqa_verified.csv) CC BY 4.0third-party · Retrieved 2026-11-12 | Run by Epoch AI (third-party measurement). Source label "Claude Haiku 5.5 (max)". effort is indicated by the suffix _max in Epoch's Model version. Data retrieved 2026-11-12.transcribed from tableSource excerptsimpleqa_verified.csv: Model version=claude-haiku-5-5_max, Best score (across scorers)=0.238, stderr=0.013473586661967093, Started at=2026-10-09T14:36:06.000Z |
| Chess Puzzles(Epoch AI)Low | Low | 24.0% | — | Epoch AI | Epoch AI, Capabilities & benchmarking (benchmark_data.zip, chess_puzzles.csv) CC BY 4.0third-party · Retrieved 2026-11-12 | Run by Epoch AI (third-party measurement). Source label "Claude Haiku 5.5 (low)". effort is indicated by the suffix _low in Epoch's Model version. Data retrieved 2026-11-12.transcribed from tableSource excerptchess_puzzles.csv: Model version=claude-haiku-5-5_low, Best score (across scorers)=0.24, stderr=0.04292346959909278, Started at=2026-10-13T07:50:03.000Z |
| Chess Puzzles(Epoch AI)Max | Max | 55.0% | — | Epoch AI | Epoch AI, Capabilities & benchmarking (benchmark_data.zip, chess_puzzles.csv) CC BY 4.0third-party · Retrieved 2026-11-12 | Run by Epoch AI (third-party measurement). Source label "Claude Haiku 5.5 (max)". effort is indicated by the suffix _max in Epoch's Model version. Data retrieved 2026-11-12.transcribed from tableSource excerptchess_puzzles.csv: Model version=claude-haiku-5-5_max, Best score (across scorers)=0.55, stderr=0.049999999999999996, Started at=2026-10-13T12:40:01.000Z |
| MirrorCodeMax | Max | 78.9% | — | Epoch AI | Epoch AI, Capabilities & benchmarking (benchmark_data.zip, mirrorcode.csv) CC BY 4.0third-party · Retrieved 2026-11-12 | Run by Epoch AI (third-party measurement). Source label "Claude Haiku 5.5 (max)". effort is indicated by the suffix _max in Epoch's Model version. Data retrieved 2026-11-12.transcribed from tableSource excerptmirrorcode.csv: Model version=claude-haiku-5-5_max, Best score (across scorers)=0.7888888888888889, stderr=0.08045286284110595, Started at=2026-10-21T16:48:32.000Z |
| Epoch Capabilities Index(ECI)not stated | not stated | 160.43 | — | Epoch AI | Epoch AI, epoch_capabilities_index/eci_scores.csv (CC BY 4.0)third-party · Retrieved 2026-11-12 | A per-model metric that combines results across multiple benchmarks and multiple efforts (not per effort). Calculated by Epoch AI.transcribed from tableSource excerpteci_scores.csv: Model=Claude Haiku 5.5, eci=160.43, eci_ci_low=158.27, eci_ci_high=162.87 |
| LMArena Text (style control, overall)High | High | 1,422 | — | LMArena | lmarena-ai/leaderboard-dataset text_style_control/latest-00000-of-00001.parquet (CC BY 4.0)third-party · Retrieved 2026-11-12 | Human head-to-head votes (Bradley-Terry). Source label "Claude Haiku 5.5 (High)" (effort is in parentheses in the label). Release date 2026-11-09.transcribed from tableSource excerptmodel_name=Claude Haiku 5.5 (High), rating=1422.40, rating_lower=1416.1, rating_upper=1428.7, vote_count=13239, rank=131, category=overall, leaderboard_publish_date=2026-11-09 |
| LMArena Text Coding (without style control)High | High | 1,454 | — | LMArena | lmarena-ai/leaderboard-dataset text/latest-00000-of-00001.parquet (CC BY 4.0)third-party · Retrieved 2026-11-12 | Human head-to-head votes (Bradley-Terry). Source label "Claude Haiku 5.5 (High)" (effort is in parentheses in the label). Release date 2026-11-09.transcribed from tableSource excerptmodel_name=Claude Haiku 5.5 (High), rating=1454.15, rating_lower=1442.8, rating_upper=1465.5, vote_count=3132, rank=96, category=coding, leaderboard_publish_date=2026-11-09 |
Related rankings
- Overall ranking (performance first)
- Terminal-Bench 4.0 ranking by metric
- Terminal-Bench-Science 0.1 ranking by metric
- FrontierCode v1.1(main set) ranking by metric
- FrontierSWE 2 ranking by metric
- GDPval-AA v2.1 ranking by metric
- AA-Briefcase v1.1(Elo) ranking by metric
- Humanity's Last Exam (with tools) ranking by metric
- Humanity's Last Exam (without tools) ranking by metric
- OSWorld 2.1(offline subset) ranking by metric
- Chartography (with tools) ranking by metric
- Chartography (without tools) ranking by metric
- HealthBench Professional (length-adjusted score) ranking by metric
- GPQA Diamond ranking by metric
- FrontierMath Tiers 1-3(v2) ranking by metric
- FrontierMath Tier 4(v2) ranking by metric
- OTIS Mock AIME 2024-2025 ranking by metric
- SimpleQA Verified ranking by metric
- Chess Puzzles(Epoch AI) ranking by metric
- MirrorCode ranking by metric
- Epoch Capabilities Index(ECI) ranking by metric
- LMArena Text (style control, overall) ranking by metric
- LMArena Text Coding (without style control) ranking by metric
- Recommendations by role