Benchmark Almanac

ContractBench (Observation Contracts)

Claude Opus 4.6 77.8% leads since May 2026
Not in model cards

Agents complete 33 API workflow tasks. Validators check whether agents preserve returned values exactly and use them while they remain valid.

The repository table on November 19, 2026, placed Claude Opus 4.6 first at 77.8% success across 99 episodes. GPT-5.2 scored 74.7%.

Released
May 2026
Kind
Function calling
Level
University
Scored
Checked automatically
Answers
Agent task
Tasks
All public
Measures
Capability
Language
English
In and out
Text → Actions
Interaction
Agentic
Environment
APIs and tools
Refresh
One fixed set
Run by
Anyone
In views
Agents
Strongest model tested
Claude Opus 4.6 #39 on the capability index
Last model added
April 2026
Scores refreshed
November 20th, 2026
20
Entries on its own leaderboard, 18 models we track

Claude Opus 4.6 leads with 77.8% among models released by April 2026 Claude Opus 4.6 leads with 77.8% among models released by April 2026

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. 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.

The newest model with a score here came out in April 2026, and nobody has tested a newer one. Among the 20 highest-ranked models on the capability index, these have no score yet:

  1. 1 Claude Opus 4.6 77.8%
  2. 2 GPT-5.2 74.7%
  3. 3 GPT-5 70.7%
  4. 4 Qwen3.5 397B-A17B 70.7%
  5. 5 Claude Sonnet 4.5 69.7%
  6. 6 Qwen3.5-27B 64.6%
  7. 7 MiniMax-M2.5 62.6%
  8. 8 MiniMax-M2.1 60.6%
  9. 9 Qwen3.5-9B 56.6%
  10. 10 MiniMax-M2 53.5%
  11. 11 Gemini 2.5 Pro (snapshot unspecified) 51.5%
  12. 12 GPT-5.1 48.5%
  13. 13 Mistral Small 4 42.4%
  14. 14 Gemma 4 26B A4B 38.4%
  15. 15 Gemma 4 31B IT 37.4%
  16. 16 Ministral 3 14B 28.3%
  17. 17 GPT-4o (snapshot unspecified) 23.2%
  18. 18 Qwen2.5-72B 23.2%
  19. 19 Gemini 2.5 Flash Sep 2025 19.2%
  20. 20 Llama 3.3 70B Instruct 7.10%

Source: Official ContractBench (Observation Contracts) leaderboard

#EntryScoreSettingDateSource
1 Claude Opus 4.6 77.8% Feb 2026 Official leaderboard
2 GPT-5.2 74.7% Dec 2025 Official leaderboard
3 GPT-5 70.7% Aug 2025 Official leaderboard
4 Qwen3.5 397B-A17B 70.7% Feb 2026 Official leaderboard
5 Claude Sonnet 4.5 69.7% Sep 2025 Official leaderboard
6 Qwen3.5-27B 64.6% Feb 2026 Official leaderboard
7 MiniMax-M2.5 62.6% Feb 2026 Official leaderboard
8 MiniMax-M2.1 60.6% Dec 2025 Official leaderboard
9 Qwen3.5-9B 56.6% Feb 2026 Official leaderboard
10 MiniMax-M2 53.5% Oct 2025 Official leaderboard
11 Gemini 2.5 Pro (snapshot unspecified) 51.5% Official leaderboard
12 GPT-5.1 48.5% Nov 2025 Official leaderboard
13 Mistral Small 4 42.4% Mar 2026 Official leaderboard
14 Gemma 4 26B A4B 38.4% Apr 2026 Official leaderboard
15 Gemma 4 31B IT 37.4% Apr 2026 Official leaderboard
16 Ministral 3 14B 28.3% Dec 2025 Official leaderboard
17 GPT-4o (snapshot unspecified) 23.2% Official leaderboard
18 Qwen2.5-72B 23.2% Sep 2024 Official leaderboard
19 Gemini 2.5 Flash Sep 2025 19.2% Sep 2025 Official leaderboard
20 Llama 3.3 70B Instruct 7.10% Dec 2024 Official leaderboard
#ModelScoreSettingDateSource
1 Claude Opus 4.6 77.8% Feb 2026 Official leaderboard
2 GPT-5.2 74.7% Dec 2025 Official leaderboard
3 GPT-5 70.7% Aug 2025 Official leaderboard
4 Qwen3.5 397B-A17B 70.7% Feb 2026 Official leaderboard
5 Claude Sonnet 4.5 69.7% Sep 2025 Official leaderboard
6 Qwen3.5-27B 64.6% Feb 2026 Official leaderboard
7 MiniMax-M2.5 62.6% Feb 2026 Official leaderboard
8 MiniMax-M2.1 60.6% Dec 2025 Official leaderboard
9 Qwen3.5-9B 56.6% Feb 2026 Official leaderboard
10 MiniMax-M2 53.5% Oct 2025 Official leaderboard
11 GPT-5.1 48.5% Nov 2025 Official leaderboard
12 Mistral Small 4 42.4% Mar 2026 Official leaderboard
13 Gemma 4 26B A4B 38.4% Apr 2026 Official leaderboard
14 Gemma 4 31B IT 37.4% Apr 2026 Official leaderboard
15 Ministral 3 14B 28.3% Dec 2025 Official leaderboard
16 Qwen2.5-72B 23.2% Sep 2024 Official leaderboard
17 Gemini 2.5 Flash Sep 2025 19.2% Sep 2025 Official leaderboard
18 Llama 3.3 70B Instruct 7.10% Dec 2024 Official leaderboard

Number one over time

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

Anthropic
Own leaderboard Not tested since May 2026

Source: Official ContractBench (Observation Contracts) 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 Best so far
0% 20% 40% 60% 80% 100% 2025 2026 Near the ceiling (85%) Claude Opus 4.6 GPT-5 Qwen2.5-72B 0% 20% 40% 60% 80% 100% 2025 2026 Near the ceiling (85%) Claude Opus 4.6 GPT-5

Source: Official ContractBench (Observation Contracts) 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 Tool use

All 265 in Tool use
  • AutomationBench In use Agent completes end-to-end business workflows inside a simulated company exposing dozens of REST APIs. Graded on resulting state. Gemini 4 Neon 56.3%
  • MCP Atlas In use Multi-step workflows over real MCP servers. Scores tool-use competency on realistic tasks. Gemini 4 Argon 90.7%
  • BFCL V4 In use Models call functions and tools, including multi-turn agentic tasks with web search and memory. Checked against expected calls and outcomes. Qwen3.7-Max 75%
  • τ³-bench Banking In use Agents act as banking customer-service reps over many turns with a simulated user, following policy. Pass^1 on the final database state. Qwen 3.8 Max 55.1%
  • Toolathlon-Verified In use 108 long-horizon tasks across 32 apps through about 600 MCP tools. Execution-based checkers inspect final environment state. GLM-5.3-Flash 78.4%
  • DeepSearchQA In use Questions that need multi-step search and synthesis. Final answers graded against references. Kimi K3 95%

Data audit

9 of 10 checks pass, 1 warning. Look first at the Newest score check. The newest came 7 months ago, in April 2026, for Gemma 4 26B A4B. Every benchmark

Catalog record 2 pass
  • Pass Task access All tasks are public.
  • Pass Confidence Checked against a primary source.
Scores 1 warning 3 pass 1 note
  • Pass Independent runs Scores from the official leaderboard (20 entries).
  • Pass Models scored 18 models we track, plus 2 other entries on its leaderboard.
  • Warning Newest score The newest came 7 months ago, in April 2026, for Gemma 4 26B A4B.
  • Note Frontier models None of the top 10 has a score. The strongest tested is Claude Opus 4.6 (#39).
  • Pass Headroom The best is 77.8%, under 85%, so it still separates models.
Leaderboard import 3 pass
  • Pass Importer We read the board on github.com: 18 rows for models we track and 2 other entries 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