Benchmark Almanac

FrontierCode 2.0

Claude Fable 5.2 50.2% leads since November 2026
Claude Fable 5.2 50.2% leads since November 2026
In use
Version 2.0

The official leaderboard hasn’t scored a model released after Gemini 4 Neon (November 2026). Labs report their own numbers for one newer model: Inkling 2 Mini at 15.1%.

Agents resolve 132 maintainer-built tasks from deliberately ambiguous prompts. Rubric checks ask whether a maintainer would merge the change.

Draws on 55% more repositories than 1.1 and has 501 blocking checks for thoroughness, against 168 in 1.1.

Made by
Cognition
Released
October 2026
Kind
Fixing real repositories
Level
Expert
Scored
Checked automatically
Answers
Code
Tasks
Kept private
Measures
Capability
Language
English
In and out
Text, Code → Code, Actions
Interaction
Agentic
Environment
Terminal
Run by
A host
Based on
FrontierCode
In views
Agents
Strongest model tested
GPT-6.2 Astra #1 on the capability index
Last in a launch post
November 2026
Last model added
November 2026
Scores refreshed
November 19th, 2026
35
Entries on its own leaderboard, 24 models we track
10
Launch posts list it, from 8 labs
4
System cards report it, from 4 labs

Claude Fable 5.2 leads with 50.2% Claude Fable 5.2 leads with 50.2%

Best score per entry from the benchmark's own leaderboard, read on November 19th, 2026. Where a model ran on its own with several settings, we keep its best. Every run is in the next chart. Agents other companies built around a model get a pale shade of that model's lab color. Gray marks blends and entries not in our model list, such as fine-tunes and older models. None of these count toward leaders. Agents built around a model, such as Devin CLI + Gemini 4 Argon, are their own entries. We don't count them as the model's score. This view shows only models we track, each at its best.

Claude Fable 5.2 leads with 50.2%, claims included

Best score per model from every source, including labs’ own numbers and claims Epoch AI collected from reports and papers. Labs pick their own settings, so a claim can beat an independent run of the same model. Striped bars are claims, and dotted bars are reruns by Epoch AI or another board.

  1. 1 Claude Fable 5.2 xhigh 50.2%
  2. 2 Claude Opus 5.5 max 47.3%
  3. 3 Gemini 4 Neon high 46.2%
  4. 4 GPT-6.2 Astra max 44.2%
  5. 5 GPT-6 Astra max 44%
  6. 6 GPT-6.2 Sol max 43.8%
  7. 7 Gemini 4 Carbon high 42.6%
  8. 8 Claude Sonnet 5.5 max 42.2%
  9. 9 GPT-6.1 Sol max 39.6%
  10. 10 Gemini 4 Hydrogen xhigh 38.9%
  11. 11 GPT-6 Sol max 34.9%
  12. 12 Claude Fable 5.1 xhigh 34%
  13. 13 Claude Haiku 5.5 max 28.9%
  14. 14 Grok 4.7 xhigh 28.9%
  15. 15 GPT-5.6 Sol max 28.9%
  16. 16 Claude Opus 5 high 28.1%
  17. 17 Devin CLI + Gemini 4 Argon high 27.6%
  18. 18 Muse Wonder xhigh 25.3%
  19. 19 SWE-2 high 24.1%
  20. 20 SWE-2 Flash max 24%
  21. 21 GPT-5.5 xhigh 23.6%
  22. 22 GPT-6 Luna max 22.7%
  23. 23 Devin CLI + Inkling 2 Small max 22.6%
  24. 24 Devin CLI + DeepSeek V4.1 Flash 1017 max 20.2%
  25. 25 Devin CLI + GLM-5.3 max 20.1%
  26. 26 Devin CLI + Kimi K3.1 Pro max 19.9%
  27. 27 Mistral Vibe + Mistral Large 4 high 19.6%
  28. 28 Devin CLI + Kimi K3 high 18.9%
  29. 29 SWE-1.7 na 15.5%
  30. 30 Claude Opus 4.8 max 15.3%
  31. 31 Devin CLI + GLM-5.3-Flash max 14.9%
  32. 32 Devin CLI + Muse Spark Open xhigh 6.46%
  33. 33 Devin CLI + Inkling-Small max 3.96%
  34. 34 Devin CLI + Beam 1.0 max 3.41%
  35. 35 SWE-1.6 na 1.34%

Source: Official FrontierCode 2.0 leaderboard

#EntryScoreSettingDateSource
1 Claude Fable 5.2 50.2% xhigh, Claude Code Nov 2026 Official leaderboard
2 Claude Opus 5.5 47.3% max, Claude Code Sep 2026 Official leaderboard
3 Gemini 4 Neon 46.2% high, antigravity-cli Nov 2026 Official leaderboard
4 GPT-6.2 Astra 44.2% max, Codex Nov 2026 Official leaderboard
5 GPT-6 Astra 44% max, Codex Sep 2026 Official leaderboard
6 GPT-6.2 Sol 43.8% max, Codex Nov 2026 Official leaderboard
7 Gemini 4 Carbon 42.6% high, antigravity-cli Oct 2026 Official leaderboard
8 Claude Sonnet 5.5 42.2% max, Claude Code Sep 2026 Official leaderboard
9 GPT-6.1 Sol 39.6% max, Codex Sep 2026 Official leaderboard
10 Gemini 4 Hydrogen 38.9% xhigh, antigravity-cli Nov 2026 Official leaderboard
11 GPT-6 Sol 34.9% max, Codex Sep 2026 Official leaderboard
12 Claude Fable 5.1 34% xhigh, Claude Code Sep 2026 Official leaderboard
13 Claude Haiku 5.5 28.9% max, Claude Code Oct 2026 Official leaderboard
14 Grok 4.7 28.9% xhigh, Grok Build Sep 2026 Official leaderboard
15 GPT-5.6 Sol 28.9% max, Codex Jul 2026 Official leaderboard
16 Claude Opus 5 28.1% high, Claude Code Jul 2026 Official leaderboard
17 Devin CLI + Gemini 4 Argon 27.6% high Sep 2026 Official leaderboard
18 Muse Wonder 25.3% xhigh, Muse Code Oct 2026 Official leaderboard
19 SWE-2 24.1% high, Devin Sep 2026 Official leaderboard
20 SWE-2 Flash 24% max, Devin Nov 2026 Official leaderboard
21 GPT-5.5 23.6% xhigh, Codex Apr 2026 Official leaderboard
22 GPT-6 Luna 22.7% max, Codex Sep 2026 Official leaderboard
23 Devin CLI + Inkling 2 Small 22.6% max Nov 2026 Official leaderboard
24 Devin CLI + DeepSeek V4.1 Flash 1017 20.2% max Oct 2026 Official leaderboard
25 Devin CLI + GLM-5.3 20.1% max Aug 2026 Official leaderboard
26 Devin CLI + Kimi K3.1 Pro 19.9% max Oct 2026 Official leaderboard
27 Mistral Vibe + Mistral Large 4 19.6% high Oct 2026 Official leaderboard
28 Devin CLI + Kimi K3 18.9% high Jul 2026 Official leaderboard
29 SWE-1.7 15.5% na, Devin CLI Jul 2026 Official leaderboard
30 Claude Opus 4.8 15.3% max, Claude Code May 2026 Official leaderboard
31 Devin CLI + GLM-5.3-Flash 14.9% max Aug 2026 Official leaderboard
32 Devin CLI + Muse Spark Open 6.46% xhigh Oct 2026 Official leaderboard
33 Devin CLI + Inkling-Small 3.96% max Jul 2026 Official leaderboard
34 Devin CLI + Beam 1.0 3.41% max Oct 2026 Official leaderboard
35 SWE-1.6 1.34% na, Devin CLI Apr 2026 Official leaderboard
#ModelScoreSettingDateSource
1 Claude Fable 5.2 50.2% xhigh, Claude Code Nov 2026 Official leaderboard
2 Claude Opus 5.5 47.3% max, Claude Code Sep 2026 Official leaderboard
3 Gemini 4 Neon 46.2% high, antigravity-cli Nov 2026 Official leaderboard
4 GPT-6.2 Astra 44.2% max, Codex Nov 2026 Official leaderboard
5 GPT-6 Astra 44% max, Codex Sep 2026 Official leaderboard
6 GPT-6.2 Sol 43.8% max, Codex Nov 2026 Official leaderboard
7 Gemini 4 Carbon 42.6% high, antigravity-cli Oct 2026 Official leaderboard
8 Claude Sonnet 5.5 42.2% max, Claude Code Sep 2026 Official leaderboard
9 GPT-6.1 Sol 39.6% max, Codex Sep 2026 Official leaderboard
10 Gemini 4 Hydrogen 38.9% xhigh, antigravity-cli Nov 2026 Official leaderboard
11 GPT-6 Sol 34.9% max, Codex Sep 2026 Official leaderboard
12 Claude Fable 5.1 34% xhigh, Claude Code Sep 2026 Official leaderboard
13 Claude Haiku 5.5 28.9% max, Claude Code Oct 2026 Official leaderboard
14 Grok 4.7 28.9% xhigh, Grok Build Sep 2026 Official leaderboard
15 GPT-5.6 Sol 28.9% max, Codex Jul 2026 Official leaderboard
16 Claude Opus 5 28.1% high, Claude Code Jul 2026 Official leaderboard
17 Muse Wonder 25.3% xhigh, Muse Code Oct 2026 Official leaderboard
18 SWE-2 24.1% high, Devin Sep 2026 Official leaderboard
19 SWE-2 Flash 24% max, Devin Nov 2026 Official leaderboard
20 GPT-5.5 23.6% xhigh, Codex Apr 2026 Official leaderboard
21 GPT-6 Luna 22.7% max, Codex Sep 2026 Official leaderboard
22 SWE-1.7 15.5% na, Devin CLI Jul 2026 Official leaderboard
23 Claude Opus 4.8 15.3% max, Claude Code May 2026 Official leaderboard
24 SWE-1.6 1.34% na, Devin CLI Apr 2026 Official leaderboard
#ModelScoreSettingDateSource
1 Claude Fable 5.2 50.2% xhigh, Claude Code Nov 2026 Official leaderboard
2 Claude Opus 5.5 47.3% max, Claude Code Sep 2026 Official leaderboard
3 Gemini 4 Neon 46.2% high, antigravity-cli Nov 2026 Official leaderboard
4 GPT-6.2 Astra 44.2% Nov 2026 Lab model card
5 GPT-6 Astra 44% max, Codex Sep 2026 Official leaderboard
6 GPT-6.2 Sol 43.8% Nov 2026 Lab model card
7 Gemini 4 Carbon 42.6% high, antigravity-cli Oct 2026 Official leaderboard
8 Claude Sonnet 5.5 42.2% max, Claude Code Sep 2026 Official leaderboard
9 GPT-6.1 Sol 39.6% max, Codex Sep 2026 Official leaderboard
10 Gemini 4 Hydrogen 38.9% xhigh, antigravity-cli Nov 2026 Official leaderboard
11 GPT-6 Sol 34.9% max, Codex Sep 2026 Official leaderboard
12 Claude Fable 5.1 34% xhigh, Claude Code Sep 2026 Official leaderboard
13 Tencent Hy4 31% Nov 2026 Lab model card
14 Claude Haiku 5.5 28.9% max, Claude Code Oct 2026 Official leaderboard
15 Grok 4.7 28.9% xhigh, Grok Build Sep 2026 Official leaderboard
16 GPT-5.6 Sol 28.9% max, Codex Jul 2026 Official leaderboard
17 Claude Opus 5 28.1% high, Claude Code Jul 2026 Official leaderboard
18 Muse Ember 27.2% max effort Nov 2026 Lab model card
19 Inkling 2 Small 25.4% Nov 2026 Lab model card
20 Muse Wonder 25.3% xhigh, Muse Code Oct 2026 Official leaderboard
21 SWE-2 24.1% high, Devin Sep 2026 Official leaderboard
22 SWE-2 Flash 24% max, Devin Nov 2026 Official leaderboard
23 GPT-5.5 23.6% xhigh, Codex Apr 2026 Official leaderboard
24 GPT-6 Luna 22.7% max, Codex Sep 2026 Official leaderboard
25 Mistral Large 4 (Le Chonk, open weights) 19.7% Oct 2026 Lab model card
26 SWE-1.7 15.5% na, Devin CLI Jul 2026 Official leaderboard
27 Claude Opus 4.8 15.3% max, Claude Code May 2026 Official leaderboard
28 Inkling 2 Mini 15.1% Nov 2026 Lab model card
29 SWE-1.6 1.34% na, Devin CLI Apr 2026 Official leaderboard

Labs report scores for 10 models in their cards. Switch the scores above to rank them together.

The best score, 50.2%, costs $18 a task

Runs on the cost frontier score higher than any run that's cheaper. Gemini 4 Neon at high comes within 3.9 points of the best for $4.38 a task, 4 times cheaper. The costs are the leaderboard's own estimates, so compare them only within this benchmark.

Anthropic OpenAI Cognition Google DeepMind Cost frontier One model at each effort level
0% 20% 40% 60% $0.01 $0.10 $1.00 $10 $100 Cost per task, log scale Claude Fable 5.2, xhigh Claude Opus 5.5, max GPT-6.2 Astra, xhigh GPT-6.2 Sol, max GPT-6.2 Sol, xhigh GPT-6.1 Sol, high Claude Haiku 5.5, max 0% 20% 40% 60% $0.01 $0.10 $1.00 $10 $100 Cost per task, log scale Claude Fable 5.2, xhigh Gemini 4 Neon, high

Source: Official FrontierCode 2.0 leaderboard

ModelEffortHarnessScoreCost per taskOutput tokens
Claude Fable 5.2 xhigh Claude Code 50.2% $18 156,009
Claude Fable 5.2 max Claude Code 48.2% $56 417,808
Claude Opus 5.5 max Claude Code 47.3% $13 201,919
Claude Fable 5.2 high Claude Code 46.2% $8.40 75,978
Gemini 4 Neon high antigravity-cli 46.2% $4.38 85,592
Gemini 4 Neon xhigh antigravity-cli 44.8% $7.67 127,992
Claude Opus 5.5 xhigh Claude Code 44.7% $5.18 88,053
GPT-6.2 Astra max Codex 44.2% $5.97 45,974
GPT-6.2 Astra xhigh Codex 44.1% $4.29 28,568
GPT-6 Astra max Codex 44% $6.68 36,354
GPT-6.2 Sol max Codex 43.8% $1.29 38,078
GPT-6.2 Astra high Codex 43.6% $3.95 24,452
Gemini 4 Carbon high antigravity-cli 42.6% $5.06 127,641
Claude Sonnet 5.5 max Claude Code 42.2% $20 514,722
GPT-6.2 Astra medium Codex 42.2% $2.47 14,364
GPT-6 Astra xhigh Codex 41.7% $4.54 20,869
Gemini 4 Carbon medium antigravity-cli 40.9% $4.49 114,385
GPT-6.2 Sol xhigh Codex 39.8% $0.98 26,387
GPT-6.2 Sol high Codex 39.7% $0.88 22,415
GPT-6.1 Sol max Codex 39.6% $1.03 41,479
Gemini 4 Hydrogen xhigh antigravity-cli 38.9% $4.95 153,739
GPT-6 Astra high Codex 38.8% $4.28 18,433
Claude Fable 5.2 medium Claude Code 38.8% $3.10 28,384
GPT-6.2 Sol medium Codex 38.3% $0.69 16,457
Gemini 4 Hydrogen high antigravity-cli 38.2% $2.40 85,996
Gemini 4 Neon medium antigravity-cli 38.2% $2.17 45,888
GPT-6.2 Astra low Codex 38.1% $1.42 8,154
GPT-6 Astra medium Codex 37.3% $3.49 14,065
Claude Sonnet 5.5 xhigh Claude Code 36.5% $3.31 80,433
Claude Opus 5.5 high Claude Code 36.3% $2.42 41,233
Claude Fable 5.2 low Claude Code 35.3% $2.62 23,907
GPT-6.1 Sol high Codex 35.1% $0.65 22,419
GPT-6 Sol max Codex 34.9% $2.35 54,181
Claude Fable 5.1 xhigh Claude Code 34% $8.68 92,813
GPT-6.2 Sol low Codex 33.1% $0.40 9,789
GPT-6.1 Sol xhigh Codex 33% $0.69 24,789
Claude Fable 5.1 max Claude Code 33% $11 119,089
Gemini 4 Carbon low antigravity-cli 31.5% $7.72 51,553
Claude Opus 5.5 medium Claude Code 31.4% $1.70 30,068
GPT-6 Astra low Codex 31.1% $2.44 8,987
GPT-6 Sol xhigh Codex 30.4% $1.44 28,587
GPT-6.1 Sol medium Codex 30.3% $0.49 15,686
Claude Haiku 5.5 max Claude Code 28.9% $0.37 205,112
Grok 4.7 xhigh Grok Build 28.9% $14 110,790
GPT-5.6 Sol max Codex 28.9% $5.50 38,352
Gemini 4 Hydrogen medium antigravity-cli 28.2% $0.90 37,841
Claude Opus 5 high Claude Code 28.1% $8.36 59,260
Devin CLI + Gemini 4 Argon high Devin CLI 27.6% $6.51 148,569
Gemini 4 Neon low antigravity-cli 27.4% $0.96 21,049
Claude Opus 5 xhigh Claude Code 27.4% $11 76,436
Claude Opus 5 max Claude Code 27.2% $13 88,196
Claude Fable 5.1 high Claude Code 26.6% $5.39 54,242
Grok 4.7 high Grok Build 26.5% $11 94,341
Grok 4.7 medium Grok Build 26.1% $8.03 72,180
GPT-6.1 Sol low Codex 25.7% $0.34 10,042
Muse Wonder xhigh Muse Code 25.3% $3.79 38,672
GPT-5.6 Sol xhigh Codex 25% $3.83 26,415
Muse Wonder high Muse Code 24.9% $3.14 32,832
GPT-5.6 Sol high Codex 24.9% $3.17 20,920
Claude Opus 5 medium Claude Code 24.6% $5.22 39,813
Devin CLI + Gemini 4 Argon medium Devin CLI 24.4% $5.69 123,879
SWE-2 high Devin 24.1% $0.96 68,546
SWE-2 Flash max Devin 24% $0.20 91,090
SWE-2 max Devin 24% $1.26 97,584
Muse Wonder medium Muse Code 23.9% $2.53 26,714
GPT-5.5 xhigh Codex 23.6% $5.28 30,925
GPT-6 Sol high Codex 23.3% $1.07 19,393
SWE-2 Flash low Devin 23.2% $0.20 95,841
Claude Sonnet 5.5 high Claude Code 22.9% $1.07 27,839
Claude Haiku 5.5 xhigh Claude Code 22.9% $0.23 127,735
GPT-6 Luna max Codex 22.7% $0.17 79,659
Devin CLI + Inkling 2 Small max Devin CLI 22.6% $0.58 95,400
Devin CLI + Gemini 4 Argon low Devin CLI 22.4% $3.98 63,291
Devin CLI + Inkling 2 Small high Devin CLI 22.3% $0.56 89,707
Claude Fable 5.1 medium Claude Code 22.3% $3.89 37,911
GPT-6 Sol medium Codex 22.1% $0.82 13,750
Muse Wonder low Muse Code 21.6% $1.84 20,301
Claude Fable 5.1 low Claude Code 21.2% $3.01 28,797
Grok 4.7 low Grok Build 21.1% $3.87 38,161
Devin CLI + Inkling 2 Small xhigh Devin CLI 20.7% $0.61 100,243
Devin CLI + DeepSeek V4.1 Flash 1017 max Devin CLI 20.2% $0.38 82,220
Devin CLI + GLM-5.3 max Devin CLI 20.1% $7.52 100,706
GPT-5.6 Sol medium Codex 20% $2.36 15,446
Devin CLI + Kimi K3.1 Pro max Devin CLI 19.9% $3.03 37,242
Devin CLI + DeepSeek V4.1 Flash 1017 high Devin CLI 19.8% $0.27 57,169
GPT-5.5 high Codex 19.7% $3.80 21,907
Mistral Vibe + Mistral Large 4 high Mistral Vibe 19.6% $1.55 88,069
Claude Opus 5 low Claude Code 19.3% $3.28 26,898
Claude Haiku 5.5 high Claude Code 19.3% $0.14 80,993
Devin CLI + Inkling 2 Small medium Devin CLI 19.1% $0.42 59,165
Devin CLI + Kimi K3 high Devin CLI 18.9% $4.47 49,854
GPT-6 Luna xhigh Codex 18.4% $0.13 50,565
Devin CLI + Kimi K3 max Devin CLI 18.4% $5.46 64,271
Devin CLI + GLM-5.3 high Devin CLI 18% $5.18 73,247
Devin CLI + Kimi K3.1 Pro high Devin CLI 17.5% $2.36 27,628
Devin CLI + Inkling 2 Small none Devin CLI 17.4% $0.44 62,628
Devin CLI + DeepSeek V4.1 Flash 1017 low Devin CLI 17.3% $0.21 40,659
GPT-6 Luna high Codex 16.5% $0.12 42,170
SWE-2 medium Devin 16.2% $0.44 28,705
GPT-6 Luna medium Codex 16.1% $0.09 31,360
Devin CLI + Inkling 2 Small low Devin CLI 15.6% $0.29 36,894
SWE-1.7 na Devin CLI 15.5% $3.52 89,251
Claude Opus 4.8 max Claude Code 15.3% $12 125,863
Claude Sonnet 5.5 medium Claude Code 15% $0.48 14,449
Devin CLI + GLM-5.3-Flash max Devin CLI 14.9% $0.79 90,044
Claude Opus 4.8 xhigh Claude Code 14.9% $9.13 88,364
Claude Opus 5.5 low Claude Code 14.8% $0.75 14,230
GPT-5.5 medium Codex 14.7% $2.95 16,320
Devin CLI + GLM-5.3 low Devin CLI 14.4% $3.19 47,199
Claude Haiku 5.5 medium Claude Code 14.1% $0.10 57,582
GPT-5.6 Sol low Codex 12.9% $1.45 9,806
Devin CLI + GLM-5.3-Flash high Devin CLI 12.9% $0.58 56,914
Devin CLI + Kimi K3.1 Pro low Devin CLI 12.9% $1.88 20,367
Devin CLI + Kimi K3 low Devin CLI 11.3% $2.98 28,740
Claude Opus 4.8 medium Claude Code 11.3% $5.59 57,192
Claude Opus 4.8 high Claude Code 11.1% $6.32 62,298
Claude Haiku 5.5 low Claude Code 10.6% $0.07 39,366
Gemini 4 Hydrogen low antigravity-cli 10.5% $0.35 14,953
GPT-6 Sol low Codex 10.5% $0.42 6,487
Claude Opus 4.8 low Claude Code 9.79% $4.20 40,687
Claude Sonnet 5.5 low Claude Code 9.66% $0.37 11,265
Mistral Vibe + Mistral Large 4 low Mistral Vibe 9.50% $0.67 28,384
Mistral Vibe + Mistral Large 4 medium Mistral Vibe 8.78% $1.03 51,238
Devin CLI + GLM-5.3-Flash low Devin CLI 8.01% $0.43 35,321
GPT-5.5 low Codex 7.32% $1.98 10,332
Devin CLI + Muse Spark Open xhigh Devin CLI 6.46% $9.16 44,615
Devin CLI + Muse Spark Open high Devin CLI 4.15% $6.78 33,810
GPT-6 Luna low Codex 4.05% $0.02 6,519
Devin CLI + Inkling-Small max Devin CLI 3.96% $3.68 39,077
Devin CLI + Beam 1.0 max Devin CLI 3.41% $1.30 27,971
Devin CLI + Beam 1.0 xhigh Devin CLI 2.69% $1.13 23,156
Devin CLI + Beam 1.0 high Devin CLI 1.99% $0.61 14,068
Devin CLI + Inkling-Small high Devin CLI 1.78% $1.37 13,658
Devin CLI + Muse Spark Open medium Devin CLI 1.63% $1.86 11,576
SWE-1.6 na Devin CLI 1.34% $1.36 23,990
Devin CLI + Beam 1.0 medium Devin CLI 0.83% $0.32 7,243
Devin CLI + Muse Spark Open low Devin CLI 0.42% $0.59 4,472
Devin CLI + Beam 1.0 minimal Devin CLI 0% $0.06 1,550
Devin CLI + Beam 1.0 low Devin CLI 0% $0.14 3,490

The best score, 50.2%, costs $18 a task

Runs on the cost frontier score higher than any run that's cheaper. Gemini 4 Neon at high comes within 3.9 points of the best for $4.38 a task, 4 times cheaper. The costs are the leaderboard's own estimates, so compare them only within this benchmark.

Anthropic OpenAI Cognition Google DeepMind Cost frontier One model at each effort level
0% 20% 40% 60% $0.01 $0.10 $1.00 $10 $100 Cost per task, log scale Claude Fable 5.2, xhigh Claude Opus 5.5, max GPT-6.2 Astra, xhigh GPT-6.2 Sol, max GPT-6.2 Sol, xhigh GPT-6.1 Sol, high Claude Haiku 5.5, max 0% 20% 40% 60% $0.01 $0.10 $1.00 $10 $100 Cost per task, log scale Claude Fable 5.2, xhigh Gemini 4 Neon, high

Source: Official FrontierCode 2.0 leaderboard

ModelEffortHarnessScoreCost per taskOutput tokens
Claude Fable 5.2 xhigh Claude Code 50.2% $18 156,009
Claude Fable 5.2 max Claude Code 48.2% $56 417,808
Claude Opus 5.5 max Claude Code 47.3% $13 201,919
Claude Fable 5.2 high Claude Code 46.2% $8.40 75,978
Gemini 4 Neon high antigravity-cli 46.2% $4.38 85,592
Gemini 4 Neon xhigh antigravity-cli 44.8% $7.67 127,992
Claude Opus 5.5 xhigh Claude Code 44.7% $5.18 88,053
GPT-6.2 Astra max Codex 44.2% $5.97 45,974
GPT-6.2 Astra xhigh Codex 44.1% $4.29 28,568
GPT-6 Astra max Codex 44% $6.68 36,354
GPT-6.2 Sol max Codex 43.8% $1.29 38,078
GPT-6.2 Astra high Codex 43.6% $3.95 24,452
Gemini 4 Carbon high antigravity-cli 42.6% $5.06 127,641
Claude Sonnet 5.5 max Claude Code 42.2% $20 514,722
GPT-6.2 Astra medium Codex 42.2% $2.47 14,364
GPT-6 Astra xhigh Codex 41.7% $4.54 20,869
Gemini 4 Carbon medium antigravity-cli 40.9% $4.49 114,385
GPT-6.2 Sol xhigh Codex 39.8% $0.98 26,387
GPT-6.2 Sol high Codex 39.7% $0.88 22,415
GPT-6.1 Sol max Codex 39.6% $1.03 41,479
Gemini 4 Hydrogen xhigh antigravity-cli 38.9% $4.95 153,739
GPT-6 Astra high Codex 38.8% $4.28 18,433
Claude Fable 5.2 medium Claude Code 38.8% $3.10 28,384
GPT-6.2 Sol medium Codex 38.3% $0.69 16,457
Gemini 4 Hydrogen high antigravity-cli 38.2% $2.40 85,996
Gemini 4 Neon medium antigravity-cli 38.2% $2.17 45,888
GPT-6.2 Astra low Codex 38.1% $1.42 8,154
GPT-6 Astra medium Codex 37.3% $3.49 14,065
Claude Sonnet 5.5 xhigh Claude Code 36.5% $3.31 80,433
Claude Opus 5.5 high Claude Code 36.3% $2.42 41,233
Claude Fable 5.2 low Claude Code 35.3% $2.62 23,907
GPT-6.1 Sol high Codex 35.1% $0.65 22,419
GPT-6 Sol max Codex 34.9% $2.35 54,181
Claude Fable 5.1 xhigh Claude Code 34% $8.68 92,813
GPT-6.2 Sol low Codex 33.1% $0.40 9,789
GPT-6.1 Sol xhigh Codex 33% $0.69 24,789
Claude Fable 5.1 max Claude Code 33% $11 119,089
Gemini 4 Carbon low antigravity-cli 31.5% $7.72 51,553
Claude Opus 5.5 medium Claude Code 31.4% $1.70 30,068
GPT-6 Astra low Codex 31.1% $2.44 8,987
GPT-6 Sol xhigh Codex 30.4% $1.44 28,587
GPT-6.1 Sol medium Codex 30.3% $0.49 15,686
Claude Haiku 5.5 max Claude Code 28.9% $0.37 205,112
Grok 4.7 xhigh Grok Build 28.9% $14 110,790
GPT-5.6 Sol max Codex 28.9% $5.50 38,352
Gemini 4 Hydrogen medium antigravity-cli 28.2% $0.90 37,841
Claude Opus 5 high Claude Code 28.1% $8.36 59,260
Gemini 4 Neon low antigravity-cli 27.4% $0.96 21,049
Claude Opus 5 xhigh Claude Code 27.4% $11 76,436
Claude Opus 5 max Claude Code 27.2% $13 88,196
Claude Fable 5.1 high Claude Code 26.6% $5.39 54,242
Grok 4.7 high Grok Build 26.5% $11 94,341
Grok 4.7 medium Grok Build 26.1% $8.03 72,180
GPT-6.1 Sol low Codex 25.7% $0.34 10,042
Muse Wonder xhigh Muse Code 25.3% $3.79 38,672
GPT-5.6 Sol xhigh Codex 25% $3.83 26,415
Muse Wonder high Muse Code 24.9% $3.14 32,832
GPT-5.6 Sol high Codex 24.9% $3.17 20,920
Claude Opus 5 medium Claude Code 24.6% $5.22 39,813
SWE-2 high Devin 24.1% $0.96 68,546
SWE-2 Flash max Devin 24% $0.20 91,090
SWE-2 max Devin 24% $1.26 97,584
Muse Wonder medium Muse Code 23.9% $2.53 26,714
GPT-5.5 xhigh Codex 23.6% $5.28 30,925
GPT-6 Sol high Codex 23.3% $1.07 19,393
SWE-2 Flash low Devin 23.2% $0.20 95,841
Claude Sonnet 5.5 high Claude Code 22.9% $1.07 27,839
Claude Haiku 5.5 xhigh Claude Code 22.9% $0.23 127,735
GPT-6 Luna max Codex 22.7% $0.17 79,659
Claude Fable 5.1 medium Claude Code 22.3% $3.89 37,911
GPT-6 Sol medium Codex 22.1% $0.82 13,750
Muse Wonder low Muse Code 21.6% $1.84 20,301
Claude Fable 5.1 low Claude Code 21.2% $3.01 28,797
Grok 4.7 low Grok Build 21.1% $3.87 38,161
GPT-5.6 Sol medium Codex 20% $2.36 15,446
GPT-5.5 high Codex 19.7% $3.80 21,907
Claude Opus 5 low Claude Code 19.3% $3.28 26,898
Claude Haiku 5.5 high Claude Code 19.3% $0.14 80,993
GPT-6 Luna xhigh Codex 18.4% $0.13 50,565
GPT-6 Luna high Codex 16.5% $0.12 42,170
SWE-2 medium Devin 16.2% $0.44 28,705
GPT-6 Luna medium Codex 16.1% $0.09 31,360
SWE-1.7 na Devin CLI 15.5% $3.52 89,251
Claude Opus 4.8 max Claude Code 15.3% $12 125,863
Claude Sonnet 5.5 medium Claude Code 15% $0.48 14,449
Claude Opus 4.8 xhigh Claude Code 14.9% $9.13 88,364
Claude Opus 5.5 low Claude Code 14.8% $0.75 14,230
GPT-5.5 medium Codex 14.7% $2.95 16,320
Claude Haiku 5.5 medium Claude Code 14.1% $0.10 57,582
GPT-5.6 Sol low Codex 12.9% $1.45 9,806
Claude Opus 4.8 medium Claude Code 11.3% $5.59 57,192
Claude Opus 4.8 high Claude Code 11.1% $6.32 62,298
Claude Haiku 5.5 low Claude Code 10.6% $0.07 39,366
Gemini 4 Hydrogen low antigravity-cli 10.5% $0.35 14,953
GPT-6 Sol low Codex 10.5% $0.42 6,487
Claude Opus 4.8 low Claude Code 9.79% $4.20 40,687
Claude Sonnet 5.5 low Claude Code 9.66% $0.37 11,265
GPT-5.5 low Codex 7.32% $1.98 10,332
GPT-6 Luna low Codex 4.05% $0.02 6,519
SWE-1.6 na Devin CLI 1.34% $1.36 23,990

2 models have held number one on the official leaderboard since its release in October 2026 2 models have held number one with claims included since its release in October 2026

Lab claims count from the day the card came out, independent runs from the model’s release. We draw one row each for the benchmark’s own leaderboard and what labs claim in their cards.

Anthropic Mistral AI Tencent
Own leaderboard 2 leaders
Lab claims 3 leaders

Sources: Official FrontierCode 2.0 leaderboard and labs’ model cards

Scores over time

All the scores we have, one dot each, so a model tested at several settings appears more than once.

Plot with other benchmarks
On the benchmark’s own leaderboard Reported by the lab in its model card Best so far
0% 20% 40% 60% 80% 100% 2026 2027 Near the ceiling (85%) Claude Fable 5.2 GPT-6 Astra Claude Fable 5.1 GPT-5.6 Sol GPT-5.5 SWE-1.6 0% 20% 40% 60% 80% 100% 2026 2027 Near the ceiling (85%) Claude Fable 5.2 Claude Fable 5.1

Sources: Official FrontierCode 2.0 leaderboard and labs’ model cards

8 labs put it in a launch post in the last six months

Launch posts from every lab that has listed it, one tick per post. Tall colored ticks include it, and short gray ones leave it out.

Mistral AI
Cognition
Tencent
Thinking Machines
Anthropic
Meta
OpenAI
Google DeepMind
DateLaunch postScoreAs the post writes it
Nov 20, 2026 Inkling 2 Mini 15.1 FrontierCode v2.0 (stirrup v0.2.1)
Nov 18, 2026 Gemini 4 Neon 46.2 FrontierCode 2.0 (Main score), at the default high thinking level
Nov 17, 2026 GPT-6.2 Sol 43.8 FrontierCode 2.0 (GPT-6.2 Sol Max)
Nov 17, 2026 GPT-6.2 Astra 44.2 FrontierCode 2.0 (GPT-6.2 Astra Max)
Nov 9, 2026 Muse Ember 27.2 FrontierCode 2.0-AA, Muse Ember max reasoning
Nov 9, 2026 Claude Fable 5.2 49.1 Agentic coding FrontierCode 2.0, Xhigh effort, best effort score (launch table and Table 8.1.A)
Nov 5, 2026 Inkling 2 Small 25.4 FrontierCode v2.0
Nov 4, 2026 Tencent Hy4 31 FrontierCode 2.0*
Nov 4, 2026 SWE-2 Flash 24 FrontierCode 2.0
Oct 29, 2026 Mistral Large 4 (Le Chonk, open weights) 19.7 Cognition FrontierCode 2.0

4 system cards report it, from 4 labs

Besides launch posts, we read the system cards, model cards and technical reports labs publish with a release. They report far more, including safety and frontier-risk tests.

DateDocumentScoreAs the document writes itSection
Nov 20, 2026 Inkling 2 Mini model card 15.1 FrontierCode v2.0 (stirrup v0.2.1) Capabilities
Nov 18, 2026 Gemini 4 Neon model card 46.2 FrontierCode 2.0 (Main score), at the default high thinking level (results table, p.4) Capabilities
Nov 9, 2026 Muse Ember technical report 27.2 FrontierCode 2.0-AA, Muse Ember max reasoning Capabilities
Nov 9, 2026 Claude Fable 5.2 system card 49.1 FrontierCode 2.0, mean@5, Xhigh effort, headline best-performing effort. Max effort 45.7%, or 44.4% if timed-out attempts score zero. High effort 45.3% (Table 8.1.A and Section 8.4) Capabilities

5 lab claims next to the leaderboard

Each score a lab reported for its own model, beside the same model's best score on the leaderboard. Gaps mostly come from different harnesses, effort levels and task subsets. 4 of them match, which usually means one side copied the other. Every benchmark.

Model Lab says Board says Gap
Claude Fable 5.2 49.1% Agentic coding FrontierCode 2.0, Xhigh effort, best effort score (launch table and Table 8.1.A) 50.2% Leaderboard, xhigh, Claude Code −1.1 pts
GPT-6.2 Sol 43.8% FrontierCode 2.0 (GPT-6.2 Sol Max) 43.8% Leaderboard, max, Codex Same
Gemini 4 Neon 46.2% FrontierCode 2.0 (Main score), at the default high thinking level 46.2% Leaderboard, high, antigravity-cli Same
GPT-6.2 Astra 44.2% FrontierCode 2.0 (GPT-6.2 Astra Max) 44.2% Leaderboard, max, Codex Same
SWE-2 Flash 24% FrontierCode 2.0 24% Leaderboard, max, Devin Same

Versions

We switch the top row to each new version when it comes out. Open it for one row per version. An old version stays in play while people still test it, so whoever first takes it past 85% gets credit for closing it out.

Anthropic
FrontierCode 2.0 4 leaders
FrontierCode Replaced by 1.1, Jul 2026
FrontierCode 1.1 Replaced by 2.0, Oct 2026
FrontierCode 2.0 2 leaders

Source: Official leaderboards

More in Coding

All 523 in Coding
  • SWE-bench Verified In use 500 human-validated GitHub issues from 12 Python repos. Agent writes a patch that must pass hidden tests. Claude Opus 4.7 83.5%
  • LiveCodeBench v6 In use Competitive programming problems collected after model cutoffs from LeetCode, AtCoder, Codeforces. Scored pass@1 on hidden tests. DeepSeek-V4-Pro 93.5%
  • SWE-bench Multilingual In use 300 GitHub-issue tasks across nine programming languages in SWE-bench format. Patch graded by repository tests. Claude Fable 5.2 96.6%
  • DeepSWE In use 113 original long-horizon engineering tasks across 91 repos and five languages, written from scratch. Hand-written verifiers accept alternative solutions. GPT-6 Astra 74.1%
  • SciCode In use Scientist-curated research coding problems split into subproblems across natural sciences. Code checked against test cases. Seed2.1 Pro 59.8%
  • Codeforces rating In use Model solves recent Codeforces contest problems. An Elo-style rating is computed from simulated contest performance. DeepSeek V4.1 Flash 3,471

Data audit

11 of 12 checks pass, 1 warning. Look first at the Confidence check. Likely. Some facts rest on secondary sources. Every benchmark

Catalog record 1 warning 1 pass
  • Pass Task access The tasks are private.
  • Warning Confidence Likely. Some facts rest on secondary sources.
Scores 5 pass
  • Pass Independent runs Scores from the official leaderboard (35 entries), launch posts (10 scores) and system cards (4 scores).
  • Pass Models scored 29 models we track, plus 11 other entries on its leaderboard.
  • Pass Newest score The newest is from November 2026, for Inkling 2 Mini.
  • Pass Frontier models 8 of the top 10 have a score, led by GPT-6.2 Astra (#1).
  • Pass Headroom The best is 50.2%, under 85%, so it still separates models.
Leaderboard import 3 pass
  • Pass Importer We read the board on cognition.com: 139 rows for models we track at the last read.
  • Pass Last read November 19th, 2026.
  • Pass Read worked The last read worked.
Consistency 2 pass
  • Pass Lab claims All 5 pairs agree within 10 points.
  • Pass Known issues No published audit found problems.

How we check