Benchmark Almanac

Envisioning Signals Benchmark

Claude Opus 4.7 86.0 leads since May 2026
Not in model cards

Models identify early signs of change for an industry brief. Scores combine verifiable evidence, specific claims, recent sources, and coverage across the requested topics.

Envisioning’s November leaderboard ranks 47 models across six briefs. The repository provides 12 briefs. Some models lack scores on individual briefs.

Made by
Envisioning
Released
May 2026
Kind
Running a business
Level
University
Scored
Graded by a model
Answers
Open-ended
Tasks
All public
Measures
Capability
Language
English
In and out
Text → Text
Interaction
Single turn
Refresh
One fixed set
Run by
Anyone
In views
Law and finance
Strongest model tested
Claude Opus 5.5 #5 on the capability index
Last model added
October 2026
Scores refreshed
November 20th, 2026
47
Entries on its own leaderboard, 42 models we track

2 models share the top score of 86.0 2 models share the top score of 86.0

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. 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 Opus 4.7 86.0
  2. 2 Claude Opus 5.5 86.0
  3. 3 Claude Sonnet 4.6 85.0
  4. 4 GPT-5.5 85.0
  5. 5 Claude Fable 5.1 85.0
  6. 6 Claude Sonnet 5.5 85.0
  7. 7 GPT-6.1 Sol 84.0
  8. 8 Claude Opus 4.6 83.0
  9. 9 GPT-5.4 83.0
  10. 10 DeepSeek-V4-Pro 83.0
  11. 11 Kimi K3.1 Pro 83.0
  12. 12 GPT-5.4 Mini 82.0
  13. 13 GLM-5.3 82.0
  14. 14 Kimi K2.5 81.0
  15. 15 GLM-5.1 81.0
  16. 16 Grok 4.7 81.0
  17. 17 GPT-6 Luna 81.0
  18. 18 Claude Haiku 5.5 81.0
  19. 19 perplexity/sonar-deep-research 80.0
  20. 20 o3 80.0
  21. 21 Qwen3-Max 80.0
  22. 22 Claude Haiku 4.5 80.0
  23. 23 Grok 4.1 Fast 80.0
  24. 24 DeepSeek-V3.2 80.0
  25. 25 GPT-5.6 Sol 80.0
  26. 26 DeepSeek V4 Pro 0813 80.0
  27. 27 Qwen3.8 Max (0902) 80.0
  28. 28 Gemini 4 Argon 80.0
  29. 29 mistralai/mistral-large-2512 79.0
  30. 30 Gemini 2.5 Pro Jun 2025 79.0
  31. 31 Gemini 3.1 Pro 79.0
  32. 32 Gemini 3.5 Flash 79.0
  33. 33 GPT-5.6 Terra 79.0
  34. 34 o4-mini 78.0
  35. 35 Gemini 3.1 Flash-Lite 78.0
  36. 36 Gemini 2.5 Flash Sep 2025 77.0
  37. 37 GLM-4.6 77.0
  38. 38 Claude Opus 4.8 77.0
  39. 39 Gemini 3.8 Flash 76.0
  40. 40 perplexity/sonar-reasoning-pro 74.0
  41. 41 GPT-4.1 mini 74.0
  42. 42 Grok 4 74.0
  43. 43 cohere/command-a 73.0
  44. 44 Llama 4 Maverick 73.0
  45. 45 amazon/nova-pro-v1 70.0
  46. 46 Phi-4 69.0
  47. 47 Reka Flash 3 69.0

Source: Official Envisioning Signals Benchmark leaderboard

Source: Official Envisioning Signals Benchmark leaderboard

#EntryScoreSettingDateSource
1 Claude Opus 4.7 86.0 Apr 2026 Official leaderboard
2 Claude Opus 5.5 86.0 Sep 2026 Official leaderboard
3 Claude Sonnet 4.6 85.0 Feb 2026 Official leaderboard
4 GPT-5.5 85.0 Apr 2026 Official leaderboard
5 Claude Fable 5.1 85.0 Sep 2026 Official leaderboard
6 Claude Sonnet 5.5 85.0 Sep 2026 Official leaderboard
7 GPT-6.1 Sol 84.0 Sep 2026 Official leaderboard
8 Claude Opus 4.6 83.0 Feb 2026 Official leaderboard
9 GPT-5.4 83.0 Mar 2026 Official leaderboard
10 DeepSeek-V4-Pro 83.0 Apr 2026 Official leaderboard
11 Kimi K3.1 Pro 83.0 Oct 2026 Official leaderboard
12 GPT-5.4 Mini 82.0 Mar 2026 Official leaderboard
13 GLM-5.3 82.0 Aug 2026 Official leaderboard
14 Kimi K2.5 81.0 Jan 2026 Official leaderboard
15 GLM-5.1 81.0 Apr 2026 Official leaderboard
16 Grok 4.7 81.0 Sep 2026 Official leaderboard
17 GPT-6 Luna 81.0 Sep 2026 Official leaderboard
18 Claude Haiku 5.5 81.0 Oct 2026 Official leaderboard
19 perplexity/sonar-deep-research 80.0 Official leaderboard
20 o3 80.0 Apr 2025 Official leaderboard
21 Qwen3-Max 80.0 Sep 2025 Official leaderboard
22 Claude Haiku 4.5 80.0 Oct 2025 Official leaderboard
23 Grok 4.1 Fast 80.0 Nov 2025 Official leaderboard
24 DeepSeek-V3.2 80.0 Dec 2025 Official leaderboard
25 GPT-5.6 Sol 80.0 Jul 2026 Official leaderboard
26 DeepSeek V4 Pro 0813 80.0 Aug 2026 Official leaderboard
27 Qwen3.8 Max (0902) 80.0 Sep 2026 Official leaderboard
28 Gemini 4 Argon 80.0 Sep 2026 Official leaderboard
29 mistralai/mistral-large-2512 79.0 Official leaderboard
30 Gemini 2.5 Pro Jun 2025 79.0 Jun 2025 Official leaderboard
31 Gemini 3.1 Pro 79.0 Feb 2026 Official leaderboard
32 Gemini 3.5 Flash 79.0 May 2026 Official leaderboard
33 GPT-5.6 Terra 79.0 Jul 2026 Official leaderboard
34 o4-mini 78.0 Apr 2025 Official leaderboard
35 Gemini 3.1 Flash-Lite 78.0 Mar 2026 Official leaderboard
36 Gemini 2.5 Flash Sep 2025 77.0 Sep 2025 Official leaderboard
37 GLM-4.6 77.0 Sep 2025 Official leaderboard
38 Claude Opus 4.8 77.0 May 2026 Official leaderboard
39 Gemini 3.8 Flash 76.0 Sep 2026 Official leaderboard
40 perplexity/sonar-reasoning-pro 74.0 Official leaderboard
41 GPT-4.1 mini 74.0 Apr 2025 Official leaderboard
42 Grok 4 74.0 Jul 2025 Official leaderboard
43 cohere/command-a 73.0 Official leaderboard
44 Llama 4 Maverick 73.0 Apr 2025 Official leaderboard
45 amazon/nova-pro-v1 70.0 Official leaderboard
46 Phi-4 69.0 Dec 2024 Official leaderboard
47 Reka Flash 3 69.0 Mar 2025 Official leaderboard
#ModelScoreSettingDateSource
1 Claude Opus 4.7 86.0 Apr 2026 Official leaderboard
2 Claude Opus 5.5 86.0 Sep 2026 Official leaderboard
3 Claude Sonnet 4.6 85.0 Feb 2026 Official leaderboard
4 GPT-5.5 85.0 Apr 2026 Official leaderboard
5 Claude Fable 5.1 85.0 Sep 2026 Official leaderboard
6 Claude Sonnet 5.5 85.0 Sep 2026 Official leaderboard
7 GPT-6.1 Sol 84.0 Sep 2026 Official leaderboard
8 Claude Opus 4.6 83.0 Feb 2026 Official leaderboard
9 GPT-5.4 83.0 Mar 2026 Official leaderboard
10 DeepSeek-V4-Pro 83.0 Apr 2026 Official leaderboard
11 Kimi K3.1 Pro 83.0 Oct 2026 Official leaderboard
12 GPT-5.4 Mini 82.0 Mar 2026 Official leaderboard
13 GLM-5.3 82.0 Aug 2026 Official leaderboard
14 Kimi K2.5 81.0 Jan 2026 Official leaderboard
15 GLM-5.1 81.0 Apr 2026 Official leaderboard
16 Grok 4.7 81.0 Sep 2026 Official leaderboard
17 GPT-6 Luna 81.0 Sep 2026 Official leaderboard
18 Claude Haiku 5.5 81.0 Oct 2026 Official leaderboard
19 o3 80.0 Apr 2025 Official leaderboard
20 Qwen3-Max 80.0 Sep 2025 Official leaderboard
21 Claude Haiku 4.5 80.0 Oct 2025 Official leaderboard
22 Grok 4.1 Fast 80.0 Nov 2025 Official leaderboard
23 DeepSeek-V3.2 80.0 Dec 2025 Official leaderboard
24 GPT-5.6 Sol 80.0 Jul 2026 Official leaderboard
25 DeepSeek V4 Pro 0813 80.0 Aug 2026 Official leaderboard
26 Qwen3.8 Max (0902) 80.0 Sep 2026 Official leaderboard
27 Gemini 4 Argon 80.0 Sep 2026 Official leaderboard
28 Gemini 2.5 Pro Jun 2025 79.0 Jun 2025 Official leaderboard
29 Gemini 3.1 Pro 79.0 Feb 2026 Official leaderboard
30 Gemini 3.5 Flash 79.0 May 2026 Official leaderboard
31 GPT-5.6 Terra 79.0 Jul 2026 Official leaderboard
32 o4-mini 78.0 Apr 2025 Official leaderboard
33 Gemini 3.1 Flash-Lite 78.0 Mar 2026 Official leaderboard
34 Gemini 2.5 Flash Sep 2025 77.0 Sep 2025 Official leaderboard
35 GLM-4.6 77.0 Sep 2025 Official leaderboard
36 Claude Opus 4.8 77.0 May 2026 Official leaderboard
37 Gemini 3.8 Flash 76.0 Sep 2026 Official leaderboard
38 GPT-4.1 mini 74.0 Apr 2025 Official leaderboard
39 Grok 4 74.0 Jul 2025 Official leaderboard
40 Llama 4 Maverick 73.0 Apr 2025 Official leaderboard
41 Phi-4 69.0 Dec 2024 Official leaderboard
42 Reka Flash 3 69.0 Mar 2025 Official leaderboard

Rankings by part

We read 6 parts from the benchmark’s own leaderboard, each ranking models on one part of the test, such as AI in regulated healthcare, Stablecoin payment rails and Grid + electrification. Claude Opus 5.5 leads 2 of them, and 4 other models lead the rest. Arrows show how far a model moved from its overall rank among the models we track.

AI in regulated healthcare

42 models we track have a score here.

  1. 1 Claude Opus 5.5 ↑1 85.0
  2. 2 GPT-5.4 ↑7 84.0
  3. 3 DeepSeek-V4-Pro ↑7 84.0
  4. 4 Qwen3-Max ↑16 83.0
  5. 5 Claude Sonnet 4.6 ↓2 83.0
  6. 6 GPT-5.4 Mini ↑6 83.0
  7. 7 GPT-5.5 ↓3 83.0
  8. 8 Claude Opus 4.7 ↓7 82.0
  9. 9 GPT-5.6 Terra ↑22 82.0
  10. 10 o4-mini ↑22 81.0

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 Envisioning Signals Benchmark 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
65.0 70.0 75.0 80.0 85.0 90.0 2025 2026 2027 65.0 70.0 75.0 80.0 85.0 90.0 2025 2026 2027

Source: Official Envisioning Signals Benchmark 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

All 10 checks pass. Every benchmark

Catalog record 2 pass
  • Pass Task access All tasks are public.
  • Pass Confidence Checked against a primary source.
Scores 4 pass 1 note
  • Pass Independent runs Scores from the official leaderboard (47 entries).
  • Pass Models scored 42 models we track, plus 5 other entries on its leaderboard.
  • Pass Newest score The newest is from October 2026, for Kimi K3.1 Pro.
  • Pass Frontier models 5 of the top 10 have a score, led by Claude Opus 5.5 (#5).
  • Note Headroom Scored as a plain number, with no fixed ceiling.
Leaderboard import 3 pass
  • Pass Importer We read the board on signals.envisioning.com: 42 rows for models we track and 5 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