Benchmark Almanac

GeoBenchX v2

Claude Opus 5.5 84% leads since September 2026
Not in model cards
Version 2

Agents use GIS tools to join data, measure spatial relationships, and draw maps. Three model judges compare their work with corrected reference solutions.

The revised release tested eight models and classified 123 of 202 tasks as solvable. The corrected task file predates the October paper.

Released
June 2025
Kind
Apps and MCP servers
Level
University
Scored
Checked automatically
Answers
Agent task
Tasks
All public
Measures
Capability
Language
English
Interaction
Agentic
Environment
APIs and tools
Items from
Generated
Run by
Anyone
Based on
GeoBenchX
Occupations
Architecture and engineering
In views
Agents
Strongest model tested
Claude Opus 5.5 #5 on the capability index
Last model added
October 2026
Scores refreshed
November 20th, 2026
50
Entries on Axis Spatial’s leaderboard, 49 models we track

Claude Opus 5.5 leads with 84% Claude Opus 5.5 leads with 84%

Best score per entry from Axis Spatial’s 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 Opus 5.5 max 84%
  2. 2 Claude Fable 5.1 medium 81%
  3. 3 Claude Sonnet 5.5 high 81%
  4. 4 GPT-6.1 Sol medium 79%
  5. 5 Claude Haiku 5.5 max 78%
  6. 6 Claude Sonnet 5 high 77%
  7. 7 Gemini 3.8 Flash low 76%
  8. 8 Claude Opus 5 high 75%
  9. 9 GPT-6 Sol high 75%
  10. 10 Gemini 3.1 Pro low 74%
  11. 11 GPT-6 Luna high 74%
  12. 12 Claude Opus 4.7 high 73%
  13. 13 Claude Opus 4.8 high 73%
  14. 14 GLM-5.3-Flash max 71%
  15. 15 GPT-5.6 Sol high 70%
  16. 16 GPT-5.6 Luna xhigh 68%
  17. 17 GLM-5.3 max 68%
  18. 18 DeepSeek-V4-Pro max 67%
  19. 19 GLM-5.2 max 67%
  20. 20 GPT-5.6 Terra high 67%
  21. 21 Qwen 3.8 27B xhigh 67%
  22. 22 DeepSeek-V4-Flash high 65%
  23. 23 Nemotron 3 Ultra high 65%
  24. 24 Kimi K3 Reasoning 65%
  25. 25 DeepSeek V4.1 Flash max 65%
  26. 26 Kimi K2.6 Reasoning 64%
  27. 27 Kimi K2.7 Code Reasoning 64%
  28. 28 MiMo-V2.6 Flash Reasoning 64%
  29. 29 Gemma 4 26B A4B Reasoning 63%
  30. 30 Gemini 3.1 Flash-Lite high 61%
  31. 31 Gemma 4 31B IT Reasoning 61%
  32. 32 Muse Glimmer high 61%
  33. 33 MiniMax-M3 Reasoning 59%
  34. 34 Mistral Large 4 high 59%
  35. 35 Claude Sonnet 4.6 high 57%
  36. 36 Nemotron 3.5 Super 120B Reasoning 56%
  37. 37 gpt-oss-120b high 55%
  38. 38 Claude Haiku 4.5 Reasoning 53%
  39. 39 Nemotron 3.5 Super VL Reasoning 53%
  40. 40 Qwen 3.6 35B-A3B Reasoning 51%
  41. 41 MiniMax-M3.1-Flash max 46%
  42. 42 glm-4.7-flash Reasoning 42%
  43. 43 gpt-oss-20b high 38%
  44. 44 Laguna XS 2.1 Reasoning 33%
  45. 45 Mistral Large 3 Non-reasoning 29%
  46. 46 Nemotron 3 Nano Omni 30B-A3B Reasoning 23%
  47. 47 Nemotron 3.5 Lightning 30B-A3B Reasoning 20%
  48. 48 Step 5 Preview Reasoning 18%
  49. 49 Llama 3.3 70B Non-reasoning 6.00%
  50. 50 Nemotron 3.5 Bolt E4B Reasoning 3.00%

Source: Axis Spatial’s leaderboard

Source: Axis Spatial’s leaderboard

#EntryScoreSettingDateLinkSource
1 Claude Opus 5.5 84% max, Claude Code Sep 2026 axisspatial.com Official leaderboard
2 Claude Fable 5.1 81% medium, Claude Code Sep 2026 axisspatial.com Official leaderboard
3 Claude Sonnet 5.5 81% high, Claude Code Sep 2026 axisspatial.com Official leaderboard
4 GPT-6.1 Sol 79% medium, Codex Sep 2026 axisspatial.com Official leaderboard
5 Claude Haiku 5.5 78% max, Claude Code Oct 2026 axisspatial.com Official leaderboard
6 Claude Sonnet 5 77% high, Claude Code Jun 2026 axisspatial.com Official leaderboard
7 Gemini 3.8 Flash 76% low, Terminus-2 Sep 2026 axisspatial.com Official leaderboard
8 Claude Opus 5 75% high, Claude Code Jul 2026 axisspatial.com Official leaderboard
9 GPT-6 Sol 75% high, Codex Sep 2026 axisspatial.com Official leaderboard
10 Gemini 3.1 Pro 74% low, Terminus-2 Feb 2026 axisspatial.com Official leaderboard
11 GPT-6 Luna 74% high, Codex Sep 2026 axisspatial.com Official leaderboard
12 Claude Opus 4.7 73% high, Claude Code Apr 2026 axisspatial.com Official leaderboard
13 Claude Opus 4.8 73% high, Claude Code May 2026 axisspatial.com Official leaderboard
14 GLM-5.3-Flash 71% max, Claude Code Aug 2026 axisspatial.com Official leaderboard
15 GPT-5.6 Sol 70% high, Codex Jul 2026 axisspatial.com Official leaderboard
16 GPT-5.6 Luna 68% xhigh, Codex Jul 2026 axisspatial.com Official leaderboard
17 GLM-5.3 68% max, Terminus-2 Aug 2026 axisspatial.com Official leaderboard
18 DeepSeek-V4-Pro 67% max, Terminus-2 Apr 2026 axisspatial.com Official leaderboard
19 GLM-5.2 67% max, Terminus-2 Jun 2026 axisspatial.com Official leaderboard
20 GPT-5.6 Terra 67% high, Codex Jul 2026 axisspatial.com Official leaderboard
21 Qwen 3.8 27B 67% xhigh, Terminus-2 Aug 2026 axisspatial.com Official leaderboard
22 DeepSeek-V4-Flash 65% high, Terminus-2 Apr 2026 axisspatial.com Official leaderboard
23 Nemotron 3 Ultra 65% high, Terminus-2 Jun 2026 axisspatial.com Official leaderboard
24 Kimi K3 65% Reasoning, Kimi Code Jul 2026 axisspatial.com Official leaderboard
25 DeepSeek V4.1 Flash 65% max, Codex Sep 2026 axisspatial.com Official leaderboard
26 Kimi K2.6 64% Reasoning, Terminus-2 Apr 2026 axisspatial.com Official leaderboard
27 Kimi K2.7 Code 64% Reasoning, Kimi Code Jun 2026 axisspatial.com Official leaderboard
28 MiMo-V2.6 Flash 64% Reasoning, MiMo Code Sep 2026 axisspatial.com Official leaderboard
29 Gemma 4 26B A4B 63% Reasoning, Terminus-2 Apr 2026 axisspatial.com Official leaderboard
30 Gemini 3.1 Flash-Lite 61% high, Terminus-2 Mar 2026 axisspatial.com Official leaderboard
31 Gemma 4 31B IT 61% Reasoning, Terminus-2 Apr 2026 axisspatial.com Official leaderboard
32 Muse Glimmer 61% high, Terminus-2 Aug 2026 axisspatial.com Official leaderboard
33 MiniMax-M3 59% Reasoning, MiniMax Code Jun 2026 axisspatial.com Official leaderboard
34 Mistral Large 4 59% high, Terminus-2 Oct 2026 axisspatial.com Official leaderboard
35 Claude Sonnet 4.6 57% high, Terminus-2 Feb 2026 axisspatial.com Official leaderboard
36 Nemotron 3.5 Super 120B 56% Reasoning, Terminus-2 axisspatial.com Official leaderboard
37 gpt-oss-120b 55% high, Codex Aug 2025 axisspatial.com Official leaderboard
38 Claude Haiku 4.5 53% Reasoning, Claude Code Oct 2025 axisspatial.com Official leaderboard
39 Nemotron 3.5 Super VL 53% Reasoning, Terminus-2 Oct 2026 axisspatial.com Official leaderboard
40 Qwen 3.6 35B-A3B 51% Reasoning, Terminus-2 Apr 2026 axisspatial.com Official leaderboard
41 MiniMax-M3.1-Flash 46% max, Terminus-2 Oct 2026 axisspatial.com Official leaderboard
42 glm-4.7-flash 42% Reasoning, Terminus-2 Jan 2026 axisspatial.com Official leaderboard
43 gpt-oss-20b 38% high, Terminus-2 Aug 2025 axisspatial.com Official leaderboard
44 Laguna XS 2.1 33% Reasoning, Terminus-2 Jul 2026 axisspatial.com Official leaderboard
45 Mistral Large 3 29% Non-reasoning, Terminus-2 Dec 2025 axisspatial.com Official leaderboard
46 Nemotron 3 Nano Omni 30B-A3B 23% Reasoning, Terminus-2 Apr 2026 axisspatial.com Official leaderboard
47 Nemotron 3.5 Lightning 30B-A3B 20% Reasoning, Terminus-2 Aug 2026 axisspatial.com Official leaderboard
48 Step 5 Preview 18% Reasoning, Terminus-2 Sep 2026 axisspatial.com Official leaderboard
49 Llama 3.3 70B 6.00% Non-reasoning, Terminus-2 Dec 2024 axisspatial.com Official leaderboard
50 Nemotron 3.5 Bolt E4B 3.00% Reasoning, Terminus-2 Oct 2026 axisspatial.com Official leaderboard
#ModelScoreSettingDateLinkSource
1 Claude Opus 5.5 84% max, Claude Code Sep 2026 axisspatial.com Official leaderboard
2 Claude Fable 5.1 81% medium, Claude Code Sep 2026 axisspatial.com Official leaderboard
3 Claude Sonnet 5.5 81% high, Claude Code Sep 2026 axisspatial.com Official leaderboard
4 GPT-6.1 Sol 79% medium, Codex Sep 2026 axisspatial.com Official leaderboard
5 Claude Haiku 5.5 78% max, Claude Code Oct 2026 axisspatial.com Official leaderboard
6 Claude Sonnet 5 77% high, Claude Code Jun 2026 axisspatial.com Official leaderboard
7 Gemini 3.8 Flash 76% low, Terminus-2 Sep 2026 axisspatial.com Official leaderboard
8 Claude Opus 5 75% high, Claude Code Jul 2026 axisspatial.com Official leaderboard
9 GPT-6 Sol 75% high, Codex Sep 2026 axisspatial.com Official leaderboard
10 Gemini 3.1 Pro 74% low, Terminus-2 Feb 2026 axisspatial.com Official leaderboard
11 GPT-6 Luna 74% high, Codex Sep 2026 axisspatial.com Official leaderboard
12 Claude Opus 4.7 73% high, Claude Code Apr 2026 axisspatial.com Official leaderboard
13 Claude Opus 4.8 73% high, Claude Code May 2026 axisspatial.com Official leaderboard
14 GLM-5.3-Flash 71% max, Claude Code Aug 2026 axisspatial.com Official leaderboard
15 GPT-5.6 Sol 70% high, Codex Jul 2026 axisspatial.com Official leaderboard
16 GPT-5.6 Luna 68% xhigh, Codex Jul 2026 axisspatial.com Official leaderboard
17 GLM-5.3 68% max, Terminus-2 Aug 2026 axisspatial.com Official leaderboard
18 DeepSeek-V4-Pro 67% max, Terminus-2 Apr 2026 axisspatial.com Official leaderboard
19 GLM-5.2 67% max, Terminus-2 Jun 2026 axisspatial.com Official leaderboard
20 GPT-5.6 Terra 67% high, Codex Jul 2026 axisspatial.com Official leaderboard
21 Qwen 3.8 27B 67% xhigh, Terminus-2 Aug 2026 axisspatial.com Official leaderboard
22 DeepSeek-V4-Flash 65% high, Terminus-2 Apr 2026 axisspatial.com Official leaderboard
23 Nemotron 3 Ultra 65% high, Terminus-2 Jun 2026 axisspatial.com Official leaderboard
24 Kimi K3 65% Reasoning, Kimi Code Jul 2026 axisspatial.com Official leaderboard
25 DeepSeek V4.1 Flash 65% max, Codex Sep 2026 axisspatial.com Official leaderboard
26 Kimi K2.6 64% Reasoning, Terminus-2 Apr 2026 axisspatial.com Official leaderboard
27 Kimi K2.7 Code 64% Reasoning, Kimi Code Jun 2026 axisspatial.com Official leaderboard
28 MiMo-V2.6 Flash 64% Reasoning, MiMo Code Sep 2026 axisspatial.com Official leaderboard
29 Gemma 4 26B A4B 63% Reasoning, Terminus-2 Apr 2026 axisspatial.com Official leaderboard
30 Gemini 3.1 Flash-Lite 61% high, Terminus-2 Mar 2026 axisspatial.com Official leaderboard
31 Gemma 4 31B IT 61% Reasoning, Terminus-2 Apr 2026 axisspatial.com Official leaderboard
32 Muse Glimmer 61% high, Terminus-2 Aug 2026 axisspatial.com Official leaderboard
33 MiniMax-M3 59% Reasoning, MiniMax Code Jun 2026 axisspatial.com Official leaderboard
34 Mistral Large 4 59% high, Terminus-2 Oct 2026 axisspatial.com Official leaderboard
35 Claude Sonnet 4.6 57% high, Terminus-2 Feb 2026 axisspatial.com Official leaderboard
36 gpt-oss-120b 55% high, Codex Aug 2025 axisspatial.com Official leaderboard
37 Claude Haiku 4.5 53% Reasoning, Claude Code Oct 2025 axisspatial.com Official leaderboard
38 Nemotron 3.5 Super VL 53% Reasoning, Terminus-2 Oct 2026 axisspatial.com Official leaderboard
39 Qwen 3.6 35B-A3B 51% Reasoning, Terminus-2 Apr 2026 axisspatial.com Official leaderboard
40 MiniMax-M3.1-Flash 46% max, Terminus-2 Oct 2026 axisspatial.com Official leaderboard
41 glm-4.7-flash 42% Reasoning, Terminus-2 Jan 2026 axisspatial.com Official leaderboard
42 gpt-oss-20b 38% high, Terminus-2 Aug 2025 axisspatial.com Official leaderboard
43 Laguna XS 2.1 33% Reasoning, Terminus-2 Jul 2026 axisspatial.com Official leaderboard
44 Mistral Large 3 29% Non-reasoning, Terminus-2 Dec 2025 axisspatial.com Official leaderboard
45 Nemotron 3 Nano Omni 30B-A3B 23% Reasoning, Terminus-2 Apr 2026 axisspatial.com Official leaderboard
46 Nemotron 3.5 Lightning 30B-A3B 20% Reasoning, Terminus-2 Aug 2026 axisspatial.com Official leaderboard
47 Step 5 Preview 18% Reasoning, Terminus-2 Sep 2026 axisspatial.com Official leaderboard
48 Llama 3.3 70B 6.00% Non-reasoning, Terminus-2 Dec 2024 axisspatial.com Official leaderboard
49 Nemotron 3.5 Bolt E4B 3.00% Reasoning, Terminus-2 Oct 2026 axisspatial.com Official leaderboard

Rankings by part

We read 2 parts from Axis Spatial’s leaderboard: Solvable tasks and Correct rejection. Each ranks models on one part of the test. Claude Opus 5.5 leads Solvable tasks and Gemini 3.8 Flash leads Correct rejection. Arrows show how far a model moved from its overall rank among the models we track.

Solvable tasks

49 models we track have a score here.

  1. 1 Claude Opus 5.5 84%
  2. 2 Claude Fable 5.1 82%
  3. 3 Claude Sonnet 5.5 82%
  4. 4 GLM-5.2 ↑15 80%
  5. 5 Claude Sonnet 5 ↑1 80%
  6. 6 GPT-5.6 Sol ↑9 78%
  7. 7 Mistral Large 4 ↑27 78%
  8. 8 GPT-5.6 Luna ↑8 77%
  9. 9 MiniMax-M3 ↑24 76%
  10. 10 GLM-5.3 ↑7 76%

The best score, 84%, costs $1.70 a task

Runs on the cost frontier score higher than any run that's cheaper. GPT-6.1 Sol at medium comes within 5 points of the best for $0.06 a task, 29 times cheaper. The costs are the leaderboard's own estimates, so compare them only within this benchmark.

Anthropic OpenAI Cost frontier One model at each effort level
0% 20% 40% 60% 80% 100% $0.001 $0.01 $0.10 $1.00 $10 Cost per task, log scale Claude Opus 5.5, max Claude Opus 5.5, xhigh Claude Sonnet 5.5, high Claude Haiku 5.5, xhigh GPT-6 Luna, high GPT-6 Luna, medium 0% 20% 40% 60% 80% 100% $0.001 $0.01 $0.10 $1.00 $10 Cost per task, log scale Claude Opus 5.5, max Claude Haiku 5.5, xhigh

Source: Axis Spatial’s leaderboard

ModelEffortHarnessScoreCost per taskOutput tokens
Claude Opus 5.5 max Claude Code 84% $1.70 39,000
Claude Opus 5.5 xhigh Claude Code 83% $0.32 5,000
Claude Fable 5.1 medium Claude Code 81% $0.42 3,000
Claude Sonnet 5.5 high Claude Code 81% $0.06 2,000
Claude Sonnet 5.5 xhigh Claude Code 81% $0.15 5,000
Claude Sonnet 5.5 max Claude Code 81% $1.48 72,000
Claude Fable 5.1 high Claude Code 80% $0.54 4,000
Claude Opus 5.5 medium Claude Code 79% $0.14 2,000
Claude Opus 5.5 high Claude Code 79% $0.17 3,000
GPT-6.1 Sol medium Codex 79% $0.06 2,000
GPT-6.1 Sol xhigh Codex 79% $0.13 6,000
Claude Haiku 5.5 max Claude Code 78% $0.14 59,000
Claude Sonnet 5 high Claude Code 77% $0.24 7,000
Claude Opus 5.5 low Claude Code 77% $0.10 1,000
Claude Sonnet 5.5 low Claude Code 77% $0.05 1,000
GPT-6.1 Sol high Codex 77% $0.08 3,000
Claude Haiku 5.5 xhigh Claude Code 77% $0.03 17,000
Gemini 3.8 Flash low Terminus-2 76% $0.12 8,000
Claude Sonnet 5.5 medium Claude Code 76% $0.06 2,000
Claude Opus 5 high Claude Code 75% $0.40 5,000
GPT-6 Sol high Codex 75% $0.12 4,000
Gemini 3.1 Pro low Terminus-2 74% $0.06 2,000
GPT-6 Luna high Codex 74% $0.005 4,000
Gemini 3.1 Pro high Terminus-2 73% $0.16 8,000
Claude Opus 4.7 high Claude Code 73% $0.34 4,000
Claude Opus 4.8 high Claude Code 73% $0.29 4,000
GPT-6 Luna xhigh Codex 73% $0.01 9,000
Claude Haiku 5.5 low Claude Code 73% $0.006 3,000
Claude Haiku 5.5 high Claude Code 72% $0.01 7,000
GLM-5.3-Flash max Claude Code 71% $0.08 22,000
Claude Haiku 5.5 medium Claude Code 71% $0.008 5,000
GPT-5.6 Sol high Codex 70% $0.32 6,000
GPT-5.6 Luna xhigh Codex 68% $0.03 9,000
GLM-5.3 max Terminus-2 68% $0.37 22,000
Gemini 3.8 Flash high Terminus-2 68% $0.31 33,000
DeepSeek-V4-Pro max Terminus-2 67% $0.32 21,000
GLM-5.2 max Terminus-2 67% $0.41 34,000
GLM-5.2 high Terminus-2 67% $0.22 23,000
GLM-5.2 max Claude Code 67% $0.40 29,000
GPT-5.6 Luna high Codex 67% $0.02 7,000
GPT-5.6 Terra high Codex 67% $0.13 4,000
Qwen 3.8 27B xhigh Terminus-2 67% $0.15 25,000
DeepSeek-V4-Flash high Terminus-2 65% $0.06 27,000
DeepSeek-V4-Pro high Terminus-2 65% $0.30 17,000
Nemotron 3 Ultra high Terminus-2 65% $0.08 7,000
Kimi K3 Reasoning Kimi Code 65% $0.35 8,000
GLM-5.3 max Claude Code 65% $0.45 12,000
DeepSeek V4.1 Flash max Codex 65% $0.08 35,000
Kimi K2.6 Reasoning Terminus-2 64% $0.07 8,000
Kimi K2.6 Reasoning Kimi Code 64% $0.19 13,000
Kimi K2.7 Code Reasoning Kimi Code 64% $0.19 11,000
Kimi K3 Reasoning Terminus-2 64% $0.21 4,000
GLM-5.3 low Terminus-2 64% $0.22 10,000
MiMo-V2.6 Flash Reasoning MiMo Code 64% $0.05 36,000
Gemma 4 26B A4B Reasoning Terminus-2 63% $0.01 12,000
Kimi K2.7 Code Reasoning Terminus-2 63% $0.11 11,000
GPT-5.6 Luna medium Codex 63% $0.01 4,000
Qwen 3.8 27B Reasoning Qwen Code 63% $0.22 25,000
DeepSeek-V4-Flash max Terminus-2 62% $0.07 29,000
GPT-5.6 Luna low Codex 62% $0.007 2,000
DeepSeek V4.1 Flash max Claude Code 62% $0.17 35,000
Gemini 3.1 Flash-Lite high Terminus-2 61% $0.02 7,000
Gemma 4 31B IT Reasoning Terminus-2 61% $0.01 5,000
Muse Glimmer high Terminus-2 61% $0.06 7,000
GLM-5.2 Non-reasoning Terminus-2 60% $0.12 5,000
MiniMax-M3 Reasoning MiniMax Code 59% $0.26 43,000
Mistral Large 4 high Terminus-2 59% $0.10 14,000
GLM-5.3-Flash low Terminus-2 58% $0.03 13,000
Claude Sonnet 4.6 high Terminus-2 57% $0.20 8,000
GLM-5.3-Flash max Terminus-2 57% $0.08 33,000
GPT-6 Luna medium Codex 57% $0.003 1,000
gpt-oss-120b high Codex 55% $0.16 17,000
DeepSeek-V4-Flash high Codex 54% $0.07 18,000
Claude Haiku 4.5 Reasoning Claude Code 53% $0.10 6,000
DeepSeek-V4-Flash high Claude Code 53% $0.09 17,000
Nemotron 3.5 Super VL Reasoning Terminus-2 53% $0.01 7,000
Gemini 3.1 Flash-Lite low Terminus-2 51% $0.01 4,000
Qwen 3.6 35B-A3B Reasoning Terminus-2 51% $0.06 13,000
GPT-6 Luna low Codex 51% $0.004 1,000
Qwen 3.6 35B-A3B Reasoning Qwen Code 49% 14,000
MiniMax-M3.1-Flash max Terminus-2 46% $0.02 24,000
gpt-oss-120b high Terminus-2 45% $0.02 10,000
Gemini 3.1 Flash-Lite minimal Terminus-2 44% $0.005 1,000
glm-4.7-flash Reasoning Terminus-2 42% $0.03 11,000
glm-4.7-flash Reasoning Claude Code 39% $0.08 15,000
gpt-oss-20b high Terminus-2 38% $0.02 34,000
gpt-oss-120b low Terminus-2 37% $0.006 2,000
Laguna XS 2.1 Reasoning Terminus-2 33% $0.06 26,000
Mistral Large 3 Non-reasoning Terminus-2 29% $0.02 4,000
Nemotron 3 Nano Omni 30B-A3B Reasoning Terminus-2 23% 21,000
Nemotron 3.5 Lightning 30B-A3B Reasoning Terminus-2 20% $0.02 10,000
Step 5 Preview Reasoning Terminus-2 18% $0.11 11,000
gpt-oss-20b high Codex 10% $0.12 26,000
Llama 3.3 70B Non-reasoning Terminus-2 6.00% $0.03 3,000
Nemotron 3.5 Bolt E4B Reasoning Terminus-2 3.00% 10,000

The best score, 84%, costs $1.70 a task

Runs on the cost frontier score higher than any run that's cheaper. GPT-6.1 Sol at medium comes within 5 points of the best for $0.06 a task, 29 times cheaper. The costs are the leaderboard's own estimates, so compare them only within this benchmark.

Anthropic OpenAI Cost frontier One model at each effort level
0% 20% 40% 60% 80% 100% $0.001 $0.01 $0.10 $1.00 $10 Cost per task, log scale Claude Opus 5.5, max Claude Opus 5.5, xhigh Claude Sonnet 5.5, high Claude Haiku 5.5, xhigh GPT-6 Luna, high GPT-6 Luna, medium 0% 20% 40% 60% 80% 100% $0.001 $0.01 $0.10 $1.00 $10 Cost per task, log scale Claude Opus 5.5, max Claude Haiku 5.5, xhigh

Source: Axis Spatial’s leaderboard

ModelEffortHarnessScoreCost per taskOutput tokens
Claude Opus 5.5 max Claude Code 84% $1.70 39,000
Claude Opus 5.5 xhigh Claude Code 83% $0.32 5,000
Claude Fable 5.1 medium Claude Code 81% $0.42 3,000
Claude Sonnet 5.5 high Claude Code 81% $0.06 2,000
Claude Sonnet 5.5 xhigh Claude Code 81% $0.15 5,000
Claude Sonnet 5.5 max Claude Code 81% $1.48 72,000
Claude Fable 5.1 high Claude Code 80% $0.54 4,000
Claude Opus 5.5 medium Claude Code 79% $0.14 2,000
Claude Opus 5.5 high Claude Code 79% $0.17 3,000
GPT-6.1 Sol medium Codex 79% $0.06 2,000
GPT-6.1 Sol xhigh Codex 79% $0.13 6,000
Claude Haiku 5.5 max Claude Code 78% $0.14 59,000
Claude Sonnet 5 high Claude Code 77% $0.24 7,000
Claude Opus 5.5 low Claude Code 77% $0.10 1,000
Claude Sonnet 5.5 low Claude Code 77% $0.05 1,000
GPT-6.1 Sol high Codex 77% $0.08 3,000
Claude Haiku 5.5 xhigh Claude Code 77% $0.03 17,000
Gemini 3.8 Flash low Terminus-2 76% $0.12 8,000
Claude Sonnet 5.5 medium Claude Code 76% $0.06 2,000
Claude Opus 5 high Claude Code 75% $0.40 5,000
GPT-6 Sol high Codex 75% $0.12 4,000
Gemini 3.1 Pro low Terminus-2 74% $0.06 2,000
GPT-6 Luna high Codex 74% $0.005 4,000
Gemini 3.1 Pro high Terminus-2 73% $0.16 8,000
Claude Opus 4.7 high Claude Code 73% $0.34 4,000
Claude Opus 4.8 high Claude Code 73% $0.29 4,000
GPT-6 Luna xhigh Codex 73% $0.01 9,000
Claude Haiku 5.5 low Claude Code 73% $0.006 3,000
Claude Haiku 5.5 high Claude Code 72% $0.01 7,000
GLM-5.3-Flash max Claude Code 71% $0.08 22,000
Claude Haiku 5.5 medium Claude Code 71% $0.008 5,000
GPT-5.6 Sol high Codex 70% $0.32 6,000
GPT-5.6 Luna xhigh Codex 68% $0.03 9,000
GLM-5.3 max Terminus-2 68% $0.37 22,000
Gemini 3.8 Flash high Terminus-2 68% $0.31 33,000
DeepSeek-V4-Pro max Terminus-2 67% $0.32 21,000
GLM-5.2 max Terminus-2 67% $0.41 34,000
GLM-5.2 high Terminus-2 67% $0.22 23,000
GLM-5.2 max Claude Code 67% $0.40 29,000
GPT-5.6 Luna high Codex 67% $0.02 7,000
GPT-5.6 Terra high Codex 67% $0.13 4,000
Qwen 3.8 27B xhigh Terminus-2 67% $0.15 25,000
DeepSeek-V4-Flash high Terminus-2 65% $0.06 27,000
DeepSeek-V4-Pro high Terminus-2 65% $0.30 17,000
Nemotron 3 Ultra high Terminus-2 65% $0.08 7,000
Kimi K3 Reasoning Kimi Code 65% $0.35 8,000
GLM-5.3 max Claude Code 65% $0.45 12,000
DeepSeek V4.1 Flash max Codex 65% $0.08 35,000
Kimi K2.6 Reasoning Terminus-2 64% $0.07 8,000
Kimi K2.6 Reasoning Kimi Code 64% $0.19 13,000
Kimi K2.7 Code Reasoning Kimi Code 64% $0.19 11,000
Kimi K3 Reasoning Terminus-2 64% $0.21 4,000
GLM-5.3 low Terminus-2 64% $0.22 10,000
MiMo-V2.6 Flash Reasoning MiMo Code 64% $0.05 36,000
Gemma 4 26B A4B Reasoning Terminus-2 63% $0.01 12,000
Kimi K2.7 Code Reasoning Terminus-2 63% $0.11 11,000
GPT-5.6 Luna medium Codex 63% $0.01 4,000
Qwen 3.8 27B Reasoning Qwen Code 63% $0.22 25,000
DeepSeek-V4-Flash max Terminus-2 62% $0.07 29,000
GPT-5.6 Luna low Codex 62% $0.007 2,000
DeepSeek V4.1 Flash max Claude Code 62% $0.17 35,000
Gemini 3.1 Flash-Lite high Terminus-2 61% $0.02 7,000
Gemma 4 31B IT Reasoning Terminus-2 61% $0.01 5,000
Muse Glimmer high Terminus-2 61% $0.06 7,000
GLM-5.2 Non-reasoning Terminus-2 60% $0.12 5,000
MiniMax-M3 Reasoning MiniMax Code 59% $0.26 43,000
Mistral Large 4 high Terminus-2 59% $0.10 14,000
GLM-5.3-Flash low Terminus-2 58% $0.03 13,000
Claude Sonnet 4.6 high Terminus-2 57% $0.20 8,000
GLM-5.3-Flash max Terminus-2 57% $0.08 33,000
GPT-6 Luna medium Codex 57% $0.003 1,000
gpt-oss-120b high Codex 55% $0.16 17,000
DeepSeek-V4-Flash high Codex 54% $0.07 18,000
Claude Haiku 4.5 Reasoning Claude Code 53% $0.10 6,000
DeepSeek-V4-Flash high Claude Code 53% $0.09 17,000
Nemotron 3.5 Super VL Reasoning Terminus-2 53% $0.01 7,000
Gemini 3.1 Flash-Lite low Terminus-2 51% $0.01 4,000
Qwen 3.6 35B-A3B Reasoning Terminus-2 51% $0.06 13,000
GPT-6 Luna low Codex 51% $0.004 1,000
Qwen 3.6 35B-A3B Reasoning Qwen Code 49% 14,000
MiniMax-M3.1-Flash max Terminus-2 46% $0.02 24,000
gpt-oss-120b high Terminus-2 45% $0.02 10,000
Gemini 3.1 Flash-Lite minimal Terminus-2 44% $0.005 1,000
glm-4.7-flash Reasoning Terminus-2 42% $0.03 11,000
glm-4.7-flash Reasoning Claude Code 39% $0.08 15,000
gpt-oss-20b high Terminus-2 38% $0.02 34,000
gpt-oss-120b low Terminus-2 37% $0.006 2,000
Laguna XS 2.1 Reasoning Terminus-2 33% $0.06 26,000
Mistral Large 3 Non-reasoning Terminus-2 29% $0.02 4,000
Nemotron 3 Nano Omni 30B-A3B Reasoning Terminus-2 23% 21,000
Nemotron 3.5 Lightning 30B-A3B Reasoning Terminus-2 20% $0.02 10,000
Step 5 Preview Reasoning Terminus-2 18% $0.11 11,000
gpt-oss-20b high Codex 10% $0.12 26,000
Llama 3.3 70B Non-reasoning Terminus-2 6.00% $0.03 3,000
Nemotron 3.5 Bolt E4B Reasoning Terminus-2 3.00% 10,000

7 models have held number one on Axis Spatial’s leaderboard since its release in June 2025

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

Meta OpenAI Anthropic Google DeepMind
Axis Spatial leaderboard 7 leaders

Source: Axis Spatial’s 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 Axis Spatial’s leaderboard Best so far
0% 20% 40% 60% 80% 100% 2025 2026 2027 Near the ceiling (85%) Claude Opus 5.5 Claude Sonnet 5 Claude Sonnet 4.6 gpt-oss-120b Llama 3.3 70B 0% 20% 40% 60% 80% 100% 2025 2026 2027 Near the ceiling (85%) Claude Opus 5.5 Claude Sonnet 5

Source: Axis Spatial’s 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.

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.

OpenAI Meta Anthropic Google DeepMind
GeoBenchX v2 8 leaders
GeoBenchX Replaced by v2, Jun 2025
GPT-4o 51.5%
GeoBenchX v2 7 leaders

Source: Official leaderboards

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

All 11 checks pass. Every benchmark

Catalog record 2 pass
  • Pass Task access All tasks are public.
  • Pass Confidence Checked against a primary source.
Scores 5 pass
  • Pass Independent runs Scores from the official leaderboard (50 entries).
  • Pass Models scored 49 models we track, plus 1 other entry on its leaderboard.
  • Pass Newest score The newest is from October 2026, for Nemotron 3.5 Bolt E4B.
  • Pass Frontier models 4 of the top 10 have a score, led by Claude Opus 5.5 (#5).
  • Pass Headroom The best is 84%, under 85%, so it still separates models.
Leaderboard import 3 pass
  • Pass Importer We read the board on Axis Spatial: 95 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