Benchmark Almanac

Databricks AI Index

Claude Fable 5.1 96.0 leads since November 2026
Not in model cards

Coding agents complete tasks at different costs. The index compares cost per task and solution quality alongside enterprise usage statistics.

The coding comparison is distinct from enterprise adoption and spending measures.

Made by
Databricks
Released
November 2026
Kind
Composite work indices
Level
Expert
Scored
Settled by events
Answers
Agent task
Measures
Efficiency
Language
English
In and out
Text, Code → Actions, Code
Interaction
Agentic
Environment
Terminal
Refresh
One fixed set
Run by
A host
In views
Agents
Strongest model tested
Claude Fable 5.2 #2 on the capability index
Last model added
November 2026
Scores refreshed
November 20th, 2026
20
Entries on its own leaderboard, 19 models we track

Claude Fable 5.1 leads with 96.0 Claude Fable 5.1 leads with 96.0

Best score per entry from the benchmark's own leaderboard, read on November 20th, 2026. Where a model ran with several settings or harnesses, we keep its best. Every run is in the next chart. Names in gray aren't in our model list: fine-tunes, agent systems, products and older models. They don't count toward leaders. This view shows only models we track, each at its best.

  1. 1 Claude Fable 5.1 96.0
  2. 2 Muse Wonder 94.9
  3. 3 GLM-5.3 93.1
  4. 4 DeepSeek V4.1 Flash 1017 93.1
  5. 5 Claude Fable 5.2 92.6
  6. 6 GPT-6 Astra 92.0
  7. 7 GPT-6.1 Sol 92.0
  8. 8 Kimi K3.1 Pro 91.4
  9. 9 Gemini 4 Carbon 90.3
  10. 10 GLM-5.3-Flash 89.1
  11. 11 Claude Opus 5.5 89.1
  12. 12 GPT-6 Luna 89.1
  13. 13 GPT-6.1 Terra 89.1
  14. 14 Claude Sonnet 5.5 88.0
  15. 15 Gemini 4 Argon 85.7
  16. 16 Claude Haiku 5.5 85.1
  17. 17 Gemini 3.8 Flash 82.9
  18. 18 Grok 4.7 82.3
  19. 19 Muse Spark Open 80.6
  20. 20 Beam-501B-A23B 77.7

Source: Official Databricks AI Index leaderboard

#EntryScoreSettingDateSource
1 Claude Fable 5.1 96.0 Claude Code Sep 2026 Official leaderboard
2 Muse Wonder 94.9 Codex Oct 2026 Official leaderboard
3 GLM-5.3 93.1 Claude Code Aug 2026 Official leaderboard
4 DeepSeek V4.1 Flash 1017 93.1 Codex Oct 2026 Official leaderboard
5 Claude Fable 5.2 92.6 Claude Code Nov 2026 Official leaderboard
6 GPT-6 Astra 92.0 Claude Code Sep 2026 Official leaderboard
7 GPT-6.1 Sol 92.0 Codex Sep 2026 Official leaderboard
8 Kimi K3.1 Pro 91.4 Claude Code Oct 2026 Official leaderboard
9 Gemini 4 Carbon 90.3 Claude Code Oct 2026 Official leaderboard
10 GLM-5.3-Flash 89.1 Claude Code Aug 2026 Official leaderboard
11 Claude Opus 5.5 89.1 Claude Code Sep 2026 Official leaderboard
12 GPT-6 Luna 89.1 Claude Code Sep 2026 Official leaderboard
13 GPT-6.1 Terra 89.1 Claude Code Oct 2026 Official leaderboard
14 Claude Sonnet 5.5 88.0 Claude Code Sep 2026 Official leaderboard
15 Gemini 4 Argon 85.7 Codex Sep 2026 Official leaderboard
16 Claude Haiku 5.5 85.1 Claude Code Oct 2026 Official leaderboard
17 Gemini 3.8 Flash 82.9 Codex Sep 2026 Official leaderboard
18 Grok 4.7 82.3 Codex Sep 2026 Official leaderboard
19 Muse Spark Open 80.6 Claude Code Oct 2026 Official leaderboard
20 Beam-501B-A23B 77.7 Claude Code Official leaderboard
#ModelScoreSettingDateSource
1 Claude Fable 5.1 96.0 Claude Code Sep 2026 Official leaderboard
2 Muse Wonder 94.9 Codex Oct 2026 Official leaderboard
3 GLM-5.3 93.1 Claude Code Aug 2026 Official leaderboard
4 DeepSeek V4.1 Flash 1017 93.1 Codex Oct 2026 Official leaderboard
5 Claude Fable 5.2 92.6 Claude Code Nov 2026 Official leaderboard
6 GPT-6 Astra 92.0 Claude Code Sep 2026 Official leaderboard
7 GPT-6.1 Sol 92.0 Codex Sep 2026 Official leaderboard
8 Kimi K3.1 Pro 91.4 Claude Code Oct 2026 Official leaderboard
9 Gemini 4 Carbon 90.3 Claude Code Oct 2026 Official leaderboard
10 GLM-5.3-Flash 89.1 Claude Code Aug 2026 Official leaderboard
11 Claude Opus 5.5 89.1 Claude Code Sep 2026 Official leaderboard
12 GPT-6 Luna 89.1 Claude Code Sep 2026 Official leaderboard
13 GPT-6.1 Terra 89.1 Claude Code Oct 2026 Official leaderboard
14 Claude Sonnet 5.5 88.0 Claude Code Sep 2026 Official leaderboard
15 Gemini 4 Argon 85.7 Codex Sep 2026 Official leaderboard
16 Claude Haiku 5.5 85.1 Claude Code Oct 2026 Official leaderboard
17 Gemini 3.8 Flash 82.9 Codex Sep 2026 Official leaderboard
18 Grok 4.7 82.3 Codex Sep 2026 Official leaderboard
19 Muse Spark Open 80.6 Claude Code Oct 2026 Official leaderboard

Cost per Task vs Quality

The line joins the runs that no other run beats on both measures. We redrew this chart from the numbers on the official leaderboard.

80% 90% 100% $0$2.5$5$7.5$10$12.5 Cost per task ($), lower is better Pass rate Claude Fable 5.1Muse WonderDeepSeek V4.1 Flash 1017GPT-6.1 SolGPT-6 Luna 80% 90% 100% $0$2.5$5$7.5$10$12.5 Cost per task ($), lower is better Pass rate Claude Fable 5.1Muse WonderDeepSeek V4.1 Flash 1017GPT-6.1 SolGPT-6 Luna
ModelSettingPass rateCost per task
Claude Fable 5.1 Claude Code 96% $7.47
Muse Wonder Codex 94.9% $1.53
DeepSeek V4.1 Flash 1017 Codex 93.1% $1.02
GLM-5.3 Claude Code 93.1% $2.4
Claude Fable 5.2 Claude Code 92.6% $5.69
GPT-6.1 Sol Codex 92% $0.76
GPT-6 Astra Claude Code 92% $3.28
Kimi K3.1 Pro Claude Code 91.4% $0.97
Gemini 4 Carbon Claude Code 90.3% $1.9
GPT-6 Luna Claude Code 89.1% $0.05
GPT-6.1 Terra Claude Code 89.1% $0.12
GLM-5.3-Flash Claude Code 89.1% $0.18
Claude Opus 5.5 Claude Code 89.1% $1.62
Claude Sonnet 5.5 Claude Code 88% $0.33
Gemini 4 Argon Codex 85.7% $4.84
Claude Haiku 5.5 Claude Code 85.1% $0.16
Gemini 3.8 Flash Codex 82.9% $11.97
Grok 4.7 Codex 82.3% $1.7
Muse Spark Open Claude Code 80.6% $0.42
Beam-501B-A23B Claude Code 77.7% $0.42

Cost per Task vs Quality

The line joins the runs that no other run beats on both measures. We redrew this chart from the numbers on the official leaderboard.

80% 90% 100% $0$2.5$5$7.5$10$12.5 Cost per task ($), lower is better Pass rate Claude Fable 5.1Muse WonderDeepSeek V4.1 Flash 1017GPT-6.1 SolGPT-6 Luna 80% 90% 100% $0$2.5$5$7.5$10$12.5 Cost per task ($), lower is better Pass rate Claude Fable 5.1Muse WonderDeepSeek V4.1 Flash 1017GPT-6.1 SolGPT-6 Luna
ModelSettingPass rateCost per task
Claude Fable 5.1 Claude Code 96% $7.47
Muse Wonder Codex 94.9% $1.53
DeepSeek V4.1 Flash 1017 Codex 93.1% $1.02
GLM-5.3 Claude Code 93.1% $2.4
Claude Fable 5.2 Claude Code 92.6% $5.69
GPT-6.1 Sol Codex 92% $0.76
GPT-6 Astra Claude Code 92% $3.28
Kimi K3.1 Pro Claude Code 91.4% $0.97
Gemini 4 Carbon Claude Code 90.3% $1.9
GPT-6 Luna Claude Code 89.1% $0.05
GPT-6.1 Terra Claude Code 89.1% $0.12
GLM-5.3-Flash Claude Code 89.1% $0.18
Claude Opus 5.5 Claude Code 89.1% $1.62
Claude Sonnet 5.5 Claude Code 88% $0.33
Gemini 4 Argon Codex 85.7% $4.84
Claude Haiku 5.5 Claude Code 85.1% $0.16
Gemini 3.8 Flash Codex 82.9% $11.97
Grok 4.7 Codex 82.3% $1.7
Muse Spark Open Claude Code 80.6% $0.42

The best score, 96.0, costs $7.47 a task

Runs on the cost frontier score higher than any run that's cheaper. GPT-6.1 Sol comes within 4 of the best for $0.76 a task, 10 times cheaper. The costs are the leaderboard's own estimates, so compare them only within this benchmark.

Anthropic OpenAI Meta DeepSeek Cost frontier One model at each effort level
0.00 20.0 40.0 60.0 80.0 100.0 $0.01 $0.10 $1.00 $10 $100 Cost per task, log scale Claude Fable 5.1 Muse Wonder GPT-6.1 Sol GPT-6 Luna 0.00 20.0 40.0 60.0 80.0 100.0 $0.01 $0.10 $1.00 $10 $100 Cost per task, log scale Claude Fable 5.1 GPT-6 Luna

Source: Official Databricks AI Index leaderboard

ModelEffortHarnessScoreCost per taskOutput tokens
Claude Fable 5.1 Claude Code 96.0 $7.47
Muse Wonder Codex 94.9 $1.53
GLM-5.3 Claude Code 93.1 $2.40
DeepSeek V4.1 Flash 1017 Codex 93.1 $1.02
Claude Fable 5.2 Claude Code 92.6 $5.69
GPT-6 Astra Claude Code 92.0 $3.28
GPT-6.1 Sol Codex 92.0 $0.76
Kimi K3.1 Pro Claude Code 91.4 $0.97
Gemini 4 Carbon Claude Code 90.3 $1.90
GLM-5.3-Flash Claude Code 89.1 $0.18
Claude Opus 5.5 Claude Code 89.1 $1.62
GPT-6 Luna Claude Code 89.1 $0.05
GPT-6.1 Terra Claude Code 89.1 $0.12
Claude Sonnet 5.5 Claude Code 88.0 $0.33
Gemini 4 Argon Codex 85.7 $4.84
Claude Haiku 5.5 Claude Code 85.1 $0.16
Gemini 3.8 Flash Codex 82.9 $12
Grok 4.7 Codex 82.3 $1.70
Muse Spark Open Claude Code 80.6 $0.42

The best score, 96.0, costs $7.47 a task

Runs on the cost frontier score higher than any run that's cheaper. GPT-6.1 Sol comes within 4 of the best for $0.76 a task, 10 times cheaper. The costs are the leaderboard's own estimates, so compare them only within this benchmark.

Anthropic OpenAI Meta DeepSeek Cost frontier One model at each effort level
0.00 20.0 40.0 60.0 80.0 100.0 $0.01 $0.10 $1.00 $10 $100 Cost per task, log scale Claude Fable 5.1 Muse Wonder GPT-6.1 Sol GPT-6 Luna 0.00 20.0 40.0 60.0 80.0 100.0 $0.01 $0.10 $1.00 $10 $100 Cost per task, log scale Claude Fable 5.1 GPT-6 Luna

Source: Official Databricks AI Index leaderboard

ModelEffortHarnessScoreCost per taskOutput tokens
Claude Fable 5.1 Claude Code 96.0 $7.47
Muse Wonder Codex 94.9 $1.53
GLM-5.3 Claude Code 93.1 $2.40
DeepSeek V4.1 Flash 1017 Codex 93.1 $1.02
Claude Fable 5.2 Claude Code 92.6 $5.69
GPT-6 Astra Claude Code 92.0 $3.28
GPT-6.1 Sol Codex 92.0 $0.76
Kimi K3.1 Pro Claude Code 91.4 $0.97
Gemini 4 Carbon Claude Code 90.3 $1.90
GLM-5.3-Flash Claude Code 89.1 $0.18
Claude Opus 5.5 Claude Code 89.1 $1.62
GPT-6 Luna Claude Code 89.1 $0.05
GPT-6.1 Terra Claude Code 89.1 $0.12
Claude Sonnet 5.5 Claude Code 88.0 $0.33
Gemini 4 Argon Codex 85.7 $4.84
Claude Haiku 5.5 Claude Code 85.1 $0.16
Gemini 3.8 Flash Codex 82.9 $12
Grok 4.7 Codex 82.3 $1.70
Muse Spark Open Claude Code 80.6 $0.42

Number one over time

Lab claims count from the day the card came out, independent runs from the model’s release.

Anthropic
Own leaderboard 1 leader

Source: Official Databricks AI Index leaderboard

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
80.0 85.0 90.0 95.0 100.0 2027 80.0 85.0 90.0 95.0 100.0 2027

Source: Official Databricks AI Index leaderboard

No lab model card reports it yet

We read the headline results of 540 launch posts from 57 labs, and none of them list this benchmark.

More in Economic work

All 210 in Economic work
  • JobBench In use 130 tasks from 35 occupations workers most want delegated. Agent works through messy reference files, graded by fact-anchored binary rubrics. GLM-5.3 61.4%
  • GDPval In use Real work tasks from 44 occupations with reference files. Experts blindly compare model deliverables with expert work. Metric is win-or-tie rate. Seed2.1 Pro 87.9%
  • GDPval-AA v2.1 In use Agents with shell and web access complete 220 GDPval gold tasks from 44 occupations. Blind pairwise comparisons give Elo ratings. Claude Opus 5.5 1,846
  • AA-Briefcase v1.1 In use Version 1.1 of Artificial Analysis's multi-hour office work evaluation, with Elo ratings fitted by Crowd-BT. Claude Opus 5.5 1,822
  • Agents' Last Exam v1.1 In use Agents complete 122 public long-horizon professional workflow tasks across 55 sub-industries. Deterministic graders check outcomes against hidden references. GPT-6.2 Astra 47.5%
  • CoWorkBench In use Agent completes long-horizon productivity tasks across computer science, finance, law and medicine. Qwen in-house scoring (method unpublished). Claude Fable 5 75.9%

Data audit

9 of 10 checks pass, 1 warning. Look first at the Task access check. We don’t know whether the tasks are public. Every benchmark

Catalog record 1 warning 1 pass
  • Warning Task access We don’t know whether the tasks are public.
  • Pass Confidence Checked against a primary source.
Scores 4 pass 1 note
  • Pass Independent runs Scores from the official leaderboard (20 entries).
  • Pass Models scored 19 models we track, plus 1 other entry on its leaderboard.
  • Pass Newest score The newest is from November 2026, for Claude Fable 5.2.
  • Pass Frontier models 9 of the top 10 have a score, led by Claude Fable 5.2 (#2).
  • Note Headroom Scored as a plain number, with no fixed ceiling.
Leaderboard import 3 pass
  • Pass Importer We read the board on developers.databricks.com: 19 rows for models we track and 1 other entry at the last read.
  • Pass Last read November 20th, 2026.
  • Pass Read worked The last read worked.
Consistency 1 pass 1 note
  • Note Lab claims No model has both a lab number and an independent run.
  • Pass Known issues No published audit found problems.

How we check