Benchmark Almanac

Coverage Cat AI Insurance Benchmark (fictional insurance tasks)

GPT-6 Astra 100% leads since November 2026
Not in model cards

Models read fictional policy documents, find missing facts, and complete insurance applications. The scorer checks source quotes, required fields, and authorized actions.

Coverage Cat reports 12 models on 240 tasks. Models wrote the tasks and answer keys, which lack independent insurance expert review.

Made by
Coverage Cat
Released
November 2026
Kind
Customer service
Level
University
Scored
Checked automatically
Answers
Agent task
Measures
Capability
Language
English
In and out
Text, Documents → Text, Actions
Interaction
Agentic
Environment
APIs and tools
Items from
Generated
Refresh
One fixed set
Run by
A host
Strongest model tested
GPT-6 Astra #4 on the capability index
Last model added
October 2026
Scores refreshed
November 20th, 2026
12
Models scored on its own leaderboard

3 models share the top score of 100%

Best score per model 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.

  1. 1 GPT-6 Astra high reasoning 100%
  2. 2 GPT-6.1 Sol high reasoning 100%
  3. 3 GPT-6.1 Terra high reasoning 100%
  4. 4 Gemini 4 Argon high reasoning 99.7%
  5. 5 Grok 4.7 high reasoning 99%
  6. 6 Claude Opus 5.5 high reasoning 99%
  7. 7 GPT-6 Luna high reasoning 98.3%
  8. 8 DeepSeek V4 Pro 0813 high reasoning, fp4 93.8%
  9. 9 Claude Sonnet 5.5 high reasoning 89.6%
  10. 10 GLM-5.3 high reasoning, fp4 77.1%
  11. 11 DeepSeek-V3.2 vendor default reasoning, fp4 76.7%
  12. 12 Kimi K3 high reasoning, mxfp4 71.5%

Source: Official Coverage Cat AI Insurance Benchmark (fictional insurance tasks) leaderboard

#ModelScoreSettingDateSource
1 GPT-6 Astra 100% high reasoning Sep 2026 Official leaderboard
2 GPT-6.1 Sol 100% high reasoning Sep 2026 Official leaderboard
3 GPT-6.1 Terra 100% high reasoning Oct 2026 Official leaderboard
4 Gemini 4 Argon 99.7% high reasoning Sep 2026 Official leaderboard
5 Grok 4.7 99% high reasoning Sep 2026 Official leaderboard
6 Claude Opus 5.5 99% high reasoning Sep 2026 Official leaderboard
7 GPT-6 Luna 98.3% high reasoning Sep 2026 Official leaderboard
8 DeepSeek V4 Pro 0813 93.8% high reasoning, fp4 Aug 2026 Official leaderboard
9 Claude Sonnet 5.5 89.6% high reasoning Sep 2026 Official leaderboard
10 GLM-5.3 77.1% high reasoning, fp4 Aug 2026 Official leaderboard
11 DeepSeek-V3.2 76.7% vendor default reasoning, fp4 Dec 2025 Official leaderboard
12 Kimi K3 71.5% high reasoning, mxfp4 Jul 2026 Official leaderboard

Rankings by part

We read 3 parts from the benchmark’s own leaderboard: Read documents, Find missing facts and Complete a workflow. Each ranks models on one part of the test. GPT-6 Astra leads every one of them. Arrows show how far a model moved from its overall rank among the models we track.

Read documents

12 models we track have a score here.

  1. 1 GPT-6 Astra 100%
  2. 2 Grok 4.7 ↑3 100%
  3. 3 Claude Sonnet 5.5 ↑6 100%
  4. 4 GPT-6.1 Sol ↓2 100%
  5. 5 GPT-6.1 Terra ↓2 100%
  6. 6 Claude Opus 5.5 99%
  7. 7 Gemini 4 Argon ↓3 99%
  8. 8 GPT-6 Luna ↓1 97.9%
  9. 9 DeepSeek V4 Pro 0813 ↓1 93.8%
  10. 10 DeepSeek-V3.2 ↑1 88.5%

Of runs with a cost, the best scores 99.7% for $15 for the full run

Runs on the cost frontier score higher than any run that's cheaper. The costs are the leaderboard's own estimates, so compare them only within this benchmark.

DeepSeek Z.ai xAI Google DeepMind Cost frontier One model at each effort level
0% 20% 40% 60% 80% 100% $1.00 $2.00 $5.00 $10 $20 $50 $100 Cost of the full run, log scale Gemini 4 Argon, high reasoning DeepSeek V4 Pro 0813, high reasoning, fp4 GLM-5.3, high reasoning, fp4 0% 20% 40% 60% 80% 100% $1.00 $10 $100 Cost of the full run, log scale Gemini 4 Argon, high reasoning Grok 4.7, high reasoning

Source: Official Coverage Cat AI Insurance Benchmark (fictional insurance tasks) leaderboard

ModelEffortHarnessScoreCost of the full runOutput tokens
GPT-6 Astra high reasoning 100%
GPT-6.1 Sol high reasoning 100%
GPT-6.1 Terra high reasoning 100%
Gemini 4 Argon high reasoning 99.7% $15
Grok 4.7 high reasoning 99% $11
Claude Opus 5.5 high reasoning 99% $23
GPT-6 Luna high reasoning 98.3%
DeepSeek V4 Pro 0813 high reasoning, fp4 93.8% $3.91
Claude Sonnet 5.5 high reasoning 89.6% $9.57
GLM-5.3 high reasoning, fp4 77.1% $3.38
DeepSeek-V3.2 vendor default reasoning, fp4 76.7% $1.35
Kimi K3 high reasoning, mxfp4 71.5% $4.20

Number one over time

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

OpenAI
Own leaderboard Passed 85% in Nov 2026, by GPT-6 Astra

Source: Official Coverage Cat AI Insurance Benchmark (fictional insurance tasks) leaderboard

DeepSeek V4 Pro 0813 on the official leaderboard passed 85% in August 2026

Models released before it came out already scored that high. 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% 2026 2027 Near the ceiling (85%) GPT-6 Astra DeepSeek V4 Pro 0813 DeepSeek-V3.2 0% 20% 40% 60% 80% 100% 2026 2027 Near the ceiling (85%) GPT-6 Astra DeepSeek V4 Pro 0813

Source: Official Coverage Cat AI Insurance Benchmark (fictional insurance tasks) 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 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 (12 entries).
  • Pass Models scored 12 models we track.
  • Pass Newest score The newest is from October 2026, for GPT-6.1 Terra.
  • Pass Frontier models 6 of the top 10 have a score, led by GPT-6 Astra (#4).
  • Note Headroom Saturated. The best is 100%, past 85% since August 2026.
Leaderboard import 3 pass
  • Pass Importer We read the board on coveragecat.com: 12 rows for models we track 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