Axis Spatial

Guided workflows

GeoAgentBench and GeoAnalystBench: GIS workflows where the task gives the steps and the model writes and runs the code. The guided workflows group of the Geospatial Intelligence Index, not part of the Geospatial Agent Index.

Ongoing evaluation: results update as runs complete. Last updated 16 October 2026.

Scores

GeoAgentBench

GeoAgentBench score

Share of 50 tasks passed, 0 to 100 · Higher is better

  • Striped: not every task attempted yet
GeoAgentBench scoreScores 0 to 100: Claude Opus 5.5 (high), Claude Code: 100; Claude Haiku 5.5 (xhigh), Claude Code: 100; Claude Opus 5 (high), Claude Code: 100; GPT-6.1 Sol (medium), Codex: 100; GPT-6.1 Sol (high), Codex: 100; GPT-6.1 Sol (xhigh), Codex: 100; GPT-6 Sol (medium), Codex: 100 (not every task attempted yet); GPT-6 Sol (high), Codex: 100; GPT-5.6 Sol (high), Codex: 100; Gemini 3.8 Flash (low), Terminus-2: 98; GLM-5.3 (max), Claude Code: 98; Claude Opus 5.5 (medium), Claude Code: 98; Claude Sonnet 5.5 (medium), Claude Code: 98; Claude Sonnet 5.5 (high), Claude Code: 98; Claude Haiku 5.5 (high), Claude Code: 98; Claude Opus 5.5 (xhigh), Claude Code: 98; Claude Sonnet 5.5 (xhigh), Claude Code: 98; Claude Sonnet 5 (high), Claude Code: 98; GPT-5.6 Luna (medium), Codex: 98; GPT-5.6 Luna (high), Codex: 98; GPT-5.6 Luna (xhigh), Codex: 98 (not every task attempted yet); Claude Sonnet 4.6 (high), Terminus-2: 96; DeepSeek V4 Flash (max), Terminus-2: 96; DeepSeek V4 Pro (max), Terminus-2: 96; DeepSeek V4 Pro (high), Terminus-2: 96; Claude Opus 4.8 (high), Claude Code: 96; Claude Opus 4.7 (high), Claude Code: 96; GPT-6 Luna (xhigh), Codex: 96; GPT-6 Luna (high), Codex: 96; GPT-5.6 Terra (high), Codex: 96; DeepSeek V4 Flash (high), Terminus-2: 94; GLM-5.2 (high), Terminus-2: 94; GLM-5.2 (max), Terminus-2: 94; GLM-5.3-Flash (low), Terminus-2: 94; GLM-5.3 (max), Terminus-2: 94; Kimi K3 (Reasoning), Terminus-2: 94; Claude Haiku 5.5 (medium), Claude Code: 94; GPT-6 Luna (medium), Codex: 94; GLM-5.2 (Non-reasoning), Terminus-2: 92; GLM-5.3 (low), Terminus-2: 92; Kimi K2.7 Code (Reasoning), Terminus-2: 92; GPT-6 Luna (low), Codex: 92; Kimi K2.6 (Reasoning), Terminus-2: 90; Gemini 3.1 Pro (preview) (high), Terminus-2: 86; Gemma 4 26B A4B (Reasoning), Terminus-2: 86; GPT-5.6 Luna (low), Codex: 86; Gemini 3.1 Pro (preview) (low), Terminus-2: 84; GLM-5.3-Flash (max), Terminus-2: 84; Claude Haiku 4.5 (Reasoning), Claude Code: 78; Gemini 3.8 Flash (high), Terminus-2: 74; gpt-oss-120b (high), Terminus-2: 70; Gemini 3.1 Flash-Lite (high), Terminus-2: 69 (not every task attempted yet); Qwen 3.8 27B (xhigh), Terminus-2: 60 (not every task attempted yet); gpt-oss-120b (low), Terminus-2: 60; gpt-oss-20b (high), Terminus-2: 53 (not every task attempted yet); Gemini 3.1 Flash-Lite (low), Terminus-2: 52; Gemini 3.1 Flash-Lite (minimal), Terminus-2: 42; GLM-4.7-Flash (Reasoning), Terminus-2: 42; Llama 3.3 70B (Non-reasoning), Terminus-2: 80255075100Claude Opus 5.5 (high)Claude Code100Claude Haiku 5.5 (xhigh)Claude Code100Claude Opus 5 (high)Claude Code100GPT-6.1 Sol (medium)Codex100GPT-6.1 Sol (high)Codex100GPT-6.1 Sol (xhigh)Codex100GPT-6 Sol (medium)Codex100GPT-6 Sol (high)Codex100GPT-5.6 Sol (high)Codex100Gemini 3.8 Flash (low)Terminus-298GLM-5.3 (max)Claude Code98Claude Opus 5.5 (medium)Claude Code98Claude Sonnet 5.5 (medium)Claude Code98Claude Sonnet 5.5 (high)Claude Code98Claude Haiku 5.5 (high)Claude Code98Claude Opus 5.5 (xhigh)Claude Code98Claude Sonnet 5.5 (xhigh)Claude Code98Claude Sonnet 5 (high)Claude Code98GPT-5.6 Luna (medium)Codex98GPT-5.6 Luna (high)Codex98GPT-5.6 Luna (xhigh)Codex98Claude Sonnet 4.6 (high)Terminus-296DeepSeek V4 Flash (max)Terminus-296DeepSeek V4 Pro (max)Terminus-296DeepSeek V4 Pro (high)Terminus-296Claude Opus 4.8 (high)Claude Code96Claude Opus 4.7 (high)Claude Code96GPT-6 Luna (xhigh)Codex96GPT-6 Luna (high)Codex96GPT-5.6 Terra (high)Codex96DeepSeek V4 Flash (high)Terminus-294GLM-5.2 (high)Terminus-294GLM-5.2 (max)Terminus-294GLM-5.3-Flash (low)Terminus-294GLM-5.3 (max)Terminus-294Kimi K3 (Reasoning)Terminus-294Claude Haiku 5.5 (medium)Claude Code94GPT-6 Luna (medium)Codex94GLM-5.2 (Non-reasoning)Terminus-292GLM-5.3 (low)Terminus-292Kimi K2.7 Code (Reasoning)Terminus-292GPT-6 Luna (low)Codex92Kimi K2.6 (Reasoning)Terminus-290Gemini 3.1 Pro (preview) (high)Terminus-286Gemma 4 26B A4B (Reasoning)Terminus-286GPT-5.6 Luna (low)Codex86Gemini 3.1 Pro (preview) (low)Terminus-284GLM-5.3-Flash (max)Terminus-284Claude Haiku 4.5 (Reasoning)Claude Code78Gemini 3.8 Flash (high)Terminus-274gpt-oss-120b (high)Terminus-270Gemini 3.1 Flash-Lite (high)Terminus-269Qwen 3.8 27B (xhigh)Terminus-260gpt-oss-120b (low)Terminus-260gpt-oss-20b (high)Terminus-253Gemini 3.1 Flash-Lite (low)Terminus-252Gemini 3.1 Flash-Lite (minimal)Terminus-242GLM-4.7-Flash (Reasoning)Terminus-242Llama 3.3 70B (Non-reasoning)Terminus-28
How to read this chart

What this metric means. The share of tasks the entry passed, 0 to 100. Tasks with no scored attempt yet are left out.

Data table: GeoAgentBench score
GeoAgentBench score
ModelDeveloperGeoAgentBench scoreStatusTasks attempted
Claude Sonnet 4.6 (high), Terminus-2Anthropic96Complete50 of 50 tasks (24 attempts not counted: provider or infrastructure error)
DeepSeek V4 Flash (max), Terminus-2DeepSeek96Complete50 of 50 tasks
DeepSeek V4 Flash (high), Terminus-2DeepSeek94Complete50 of 50 tasks
DeepSeek V4 Pro (max), Terminus-2DeepSeek96Complete50 of 50 tasks
DeepSeek V4 Pro (high), Terminus-2DeepSeek96Complete50 of 50 tasks
Gemini 3.1 Flash-Lite (low), Terminus-2Google52Complete50 of 50 tasks
Gemini 3.1 Flash-Lite (minimal), Terminus-2Google42Complete50 of 50 tasks
Gemini 3.1 Flash-Lite (high), Terminus-2Google69Still to run: 2 of 50 tasks48 of 50 tasks (2 attempts not counted: provider or infrastructure error)
Gemini 3.1 Pro (preview) (low), Terminus-2Google84Complete50 of 50 tasks (4 attempts not counted: provider or infrastructure error)
Gemini 3.1 Pro (preview) (high), Terminus-2Google86Complete50 of 50 tasks
Gemini 3.8 Flash (high), Terminus-2Google74Complete50 of 50 tasks (1 attempt not counted: provider or infrastructure error)
Gemini 3.8 Flash (low), Terminus-2Google98Complete50 of 50 tasks (17 attempts not counted: provider or infrastructure error)
Gemma 4 26B A4B (Reasoning), Terminus-2Google86Complete50 of 50 tasks
GLM-4.7-Flash (Reasoning), Terminus-2Z.ai42Complete50 of 50 tasks
GLM-5.2 (high), Terminus-2Z.ai94Complete50 of 50 tasks
GLM-5.2 (Non-reasoning), Terminus-2Z.ai92Complete50 of 50 tasks
GLM-5.2 (max), Terminus-2Z.ai94Complete50 of 50 tasks
GLM-5.3-Flash (low), Terminus-2Z.ai94Complete50 of 50 tasks
GLM-5.3-Flash (max), Terminus-2Z.ai84Complete50 of 50 tasks
GLM-5.3 (low), Terminus-2Z.ai92Complete50 of 50 tasks
GLM-5.3 (max), Terminus-2Z.ai94Complete50 of 50 tasks
gpt-oss-120b (low), Terminus-2OpenAI60Complete50 of 50 tasks
gpt-oss-120b (high), Terminus-2OpenAI70Complete50 of 50 tasks
gpt-oss-20b (high), Terminus-2OpenAI53Still to run: 1 of 50 tasks49 of 50 tasks (1 attempt not counted: provider or infrastructure error)
Kimi K2.6 (Reasoning), Terminus-2Moonshot AI90Complete50 of 50 tasks
Kimi K2.7 Code (Reasoning), Terminus-2Moonshot AI92Complete50 of 50 tasks
Kimi K3 (Reasoning), Terminus-2Moonshot AI94Complete50 of 50 tasks
Llama 3.3 70B (Non-reasoning), Terminus-2Meta8Complete50 of 50 tasks
Qwen 3.8 27B (xhigh), Terminus-2Alibaba (Qwen)60Still to run: 2 of 50 tasks48 of 50 tasks
Claude Haiku 4.5 (Reasoning), Claude CodeAnthropic78Complete50 of 50 tasks
GLM-5.3 (max), Claude CodeZ.ai98Complete50 of 50 tasks
Claude Opus 5.5 (medium), Claude CodeAnthropic98Complete50 of 50 tasks
Claude Opus 5.5 (high), Claude CodeAnthropic100Complete50 of 50 tasks
Claude Sonnet 5.5 (medium), Claude CodeAnthropic98Complete50 of 50 tasks
Claude Sonnet 5.5 (high), Claude CodeAnthropic98Complete50 of 50 tasks
Claude Haiku 5.5 (medium), Claude CodeAnthropic94Complete50 of 50 tasks
Claude Haiku 5.5 (high), Claude CodeAnthropic98Complete50 of 50 tasks
Claude Opus 5.5 (xhigh), Claude CodeAnthropic98Complete50 of 50 tasks
Claude Sonnet 5.5 (xhigh), Claude CodeAnthropic98Complete50 of 50 tasks
Claude Haiku 5.5 (xhigh), Claude CodeAnthropic100Complete50 of 50 tasks
Claude Opus 5 (high), Claude CodeAnthropic100Complete50 of 50 tasks
Claude Sonnet 5 (high), Claude CodeAnthropic98Complete50 of 50 tasks
Claude Opus 4.8 (high), Claude CodeAnthropic96Complete50 of 50 tasks
Claude Opus 4.7 (high), Claude CodeAnthropic96Complete50 of 50 tasks
GPT-6.1 Sol (medium), CodexOpenAI100Complete50 of 50 tasks
GPT-6.1 Sol (high), CodexOpenAI100Complete50 of 50 tasks
GPT-6.1 Sol (xhigh), CodexOpenAI100Complete50 of 50 tasks
GPT-6 Luna (xhigh), CodexOpenAI96Complete50 of 50 tasks
GPT-6 Sol (medium), CodexOpenAI100Still to run: 35 of 50 tasks15 of 50 tasks
GPT-6 Sol (high), CodexOpenAI100Complete50 of 50 tasks
GPT-6 Luna (medium), CodexOpenAI94Complete50 of 50 tasks
GPT-6 Luna (high), CodexOpenAI96Complete50 of 50 tasks
GPT-5.6 Sol (high), CodexOpenAI100Complete50 of 50 tasks
GPT-5.6 Terra (high), CodexOpenAI96Complete50 of 50 tasks
GPT-5.6 Luna (medium), CodexOpenAI98Complete50 of 50 tasks
GPT-5.6 Luna (high), CodexOpenAI98Complete50 of 50 tasks
GPT-5.6 Luna (xhigh), CodexOpenAI98Still to run: 1 of 50 tasks49 of 50 tasks
GPT-6 Luna (low), CodexOpenAI92Complete50 of 50 tasks
GPT-5.6 Luna (low), CodexOpenAI86Complete50 of 50 tasks

GeoAnalystBench

GeoAnalystBench score

Share of 19 tasks passed, 0 to 100 · Higher is better

  • Striped: not every task attempted yet
GeoAnalystBench scoreScores 0 to 100: Claude Sonnet 4.6 (high), Terminus-2: 100; DeepSeek V4 Pro (max), Terminus-2: 100; DeepSeek V4 Pro (high), Terminus-2: 100; GLM-5.2 (high), Terminus-2: 100; Qwen 3.8 27B (xhigh), Terminus-2: 100 (not every task attempted yet); Claude Opus 5.5 (high), Claude Code: 100; Claude Sonnet 5.5 (high), Claude Code: 100; Claude Opus 5.5 (xhigh), Claude Code: 100; Claude Opus 4.8 (high), Claude Code: 100; Claude Opus 4.7 (high), Claude Code: 100; GPT-6.1 Sol (medium), Codex: 100; GPT-6.1 Sol (high), Codex: 100; GPT-6 Luna (xhigh), Codex: 100; GPT-6 Sol (high), Codex: 100; GPT-6 Luna (high), Codex: 100; GPT-5.6 Luna (medium), Codex: 100; GPT-5.6 Luna (high), Codex: 100; GPT-5.6 Luna (xhigh), Codex: 100 (not every task attempted yet); DeepSeek V4 Flash (max), Terminus-2: 95; Gemini 3.1 Pro (preview) (high), Terminus-2: 95; GLM-5.2 (max), Terminus-2: 95; GLM-5.3-Flash (low), Terminus-2: 95; GLM-5.3 (low), Terminus-2: 95; GLM-5.3 (max), Claude Code: 95; Claude Opus 5.5 (medium), Claude Code: 95; Claude Sonnet 5.5 (medium), Claude Code: 95; Claude Haiku 5.5 (medium), Claude Code: 95; Claude Sonnet 5.5 (xhigh), Claude Code: 95; Claude Haiku 5.5 (xhigh), Claude Code: 95; Claude Sonnet 5 (high), Claude Code: 95; GPT-6.1 Sol (xhigh), Codex: 95; GPT-5.6 Terra (high), Codex: 95; GPT-5.6 Luna (low), Codex: 95; DeepSeek V4 Flash (high), Terminus-2: 89; Gemini 3.8 Flash (low), Terminus-2: 89; Kimi K2.7 Code (Reasoning), Terminus-2: 89; Kimi K3 (Reasoning), Terminus-2: 89; Claude Opus 5 (high), Claude Code: 89; GPT-6 Luna (medium), Codex: 89; GPT-5.6 Sol (high), Codex: 89; GLM-5.3-Flash (max), Terminus-2: 84; GLM-5.3 (max), Terminus-2: 84; Claude Haiku 5.5 (high), Claude Code: 84; Gemini 3.1 Pro (preview) (low), Terminus-2: 79; Kimi K2.6 (Reasoning), Terminus-2: 79; GPT-6 Luna (low), Codex: 79; GLM-5.2 (Non-reasoning), Terminus-2: 74; Claude Haiku 4.5 (Reasoning), Claude Code: 74; Gemini 3.1 Flash-Lite (low), Terminus-2: 63; Gemma 4 26B A4B (Reasoning), Terminus-2: 63; Gemini 3.1 Flash-Lite (high), Terminus-2: 58; gpt-oss-120b (high), Terminus-2: 53; gpt-oss-120b (low), Terminus-2: 47; Gemini 3.1 Flash-Lite (minimal), Terminus-2: 42; GLM-4.7-Flash (Reasoning), Terminus-2: 37; gpt-oss-20b (high), Terminus-2: 37; Gemini 3.8 Flash (high), Terminus-2: 32; Llama 3.3 70B (Non-reasoning), Terminus-2: 00255075100Claude Sonnet 4.6 (high)Terminus-2100DeepSeek V4 Pro (max)Terminus-2100DeepSeek V4 Pro (high)Terminus-2100GLM-5.2 (high)Terminus-2100Qwen 3.8 27B (xhigh)Terminus-2100Claude Opus 5.5 (high)Claude Code100Claude Sonnet 5.5 (high)Claude Code100Claude Opus 5.5 (xhigh)Claude Code100Claude Opus 4.8 (high)Claude Code100Claude Opus 4.7 (high)Claude Code100GPT-6.1 Sol (medium)Codex100GPT-6.1 Sol (high)Codex100GPT-6 Luna (xhigh)Codex100GPT-6 Sol (high)Codex100GPT-6 Luna (high)Codex100GPT-5.6 Luna (medium)Codex100GPT-5.6 Luna (high)Codex100GPT-5.6 Luna (xhigh)Codex100DeepSeek V4 Flash (max)Terminus-295Gemini 3.1 Pro (preview) (high)Terminus-295GLM-5.2 (max)Terminus-295GLM-5.3-Flash (low)Terminus-295GLM-5.3 (low)Terminus-295GLM-5.3 (max)Claude Code95Claude Opus 5.5 (medium)Claude Code95Claude Sonnet 5.5 (medium)Claude Code95Claude Haiku 5.5 (medium)Claude Code95Claude Sonnet 5.5 (xhigh)Claude Code95Claude Haiku 5.5 (xhigh)Claude Code95Claude Sonnet 5 (high)Claude Code95GPT-6.1 Sol (xhigh)Codex95GPT-5.6 Terra (high)Codex95GPT-5.6 Luna (low)Codex95DeepSeek V4 Flash (high)Terminus-289Gemini 3.8 Flash (low)Terminus-289Kimi K2.7 Code (Reasoning)Terminus-289Kimi K3 (Reasoning)Terminus-289Claude Opus 5 (high)Claude Code89GPT-6 Luna (medium)Codex89GPT-5.6 Sol (high)Codex89GLM-5.3-Flash (max)Terminus-284GLM-5.3 (max)Terminus-284Claude Haiku 5.5 (high)Claude Code84Gemini 3.1 Pro (preview) (low)Terminus-279Kimi K2.6 (Reasoning)Terminus-279GPT-6 Luna (low)Codex79GLM-5.2 (Non-reasoning)Terminus-274Claude Haiku 4.5 (Reasoning)Claude Code74Gemini 3.1 Flash-Lite (low)Terminus-263Gemma 4 26B A4B (Reasoning)Terminus-263Gemini 3.1 Flash-Lite (high)Terminus-258gpt-oss-120b (high)Terminus-253gpt-oss-120b (low)Terminus-247Gemini 3.1 Flash-Lite (minimal)Terminus-242GLM-4.7-Flash (Reasoning)Terminus-237gpt-oss-20b (high)Terminus-237Gemini 3.8 Flash (high)Terminus-232Llama 3.3 70B (Non-reasoning)Terminus-20
How to read this chart

What this metric means. The share of tasks the entry passed, 0 to 100. Tasks with no scored attempt yet are left out.

Data table: GeoAnalystBench score
GeoAnalystBench score
ModelDeveloperGeoAnalystBench scoreStatusTasks attempted
Claude Sonnet 4.6 (high), Terminus-2Anthropic100Complete19 of 19 tasks (6 attempts not counted: provider or infrastructure error)
DeepSeek V4 Flash (max), Terminus-2DeepSeek95Complete19 of 19 tasks
DeepSeek V4 Flash (high), Terminus-2DeepSeek89Complete19 of 19 tasks
DeepSeek V4 Pro (max), Terminus-2DeepSeek100Complete19 of 19 tasks
DeepSeek V4 Pro (high), Terminus-2DeepSeek100Complete19 of 19 tasks
Gemini 3.1 Flash-Lite (low), Terminus-2Google63Complete19 of 19 tasks
Gemini 3.1 Flash-Lite (minimal), Terminus-2Google42Complete19 of 19 tasks
Gemini 3.1 Flash-Lite (high), Terminus-2Google58Complete19 of 19 tasks
Gemini 3.1 Pro (preview) (low), Terminus-2Google79Complete19 of 19 tasks
Gemini 3.1 Pro (preview) (high), Terminus-2Google95Complete19 of 19 tasks
Gemini 3.8 Flash (high), Terminus-2Google32Complete19 of 19 tasks
Gemini 3.8 Flash (low), Terminus-2Google89Complete19 of 19 tasks (3 attempts not counted: provider or infrastructure error)
Gemma 4 26B A4B (Reasoning), Terminus-2Google63Complete19 of 19 tasks
GLM-4.7-Flash (Reasoning), Terminus-2Z.ai37Complete19 of 19 tasks
GLM-5.2 (high), Terminus-2Z.ai100Complete19 of 19 tasks
GLM-5.2 (Non-reasoning), Terminus-2Z.ai74Complete19 of 19 tasks
GLM-5.2 (max), Terminus-2Z.ai95Complete19 of 19 tasks
GLM-5.3-Flash (low), Terminus-2Z.ai95Complete19 of 19 tasks
GLM-5.3-Flash (max), Terminus-2Z.ai84Complete19 of 19 tasks
GLM-5.3 (low), Terminus-2Z.ai95Complete19 of 19 tasks
GLM-5.3 (max), Terminus-2Z.ai84Complete19 of 19 tasks
gpt-oss-120b (low), Terminus-2OpenAI47Complete19 of 19 tasks
gpt-oss-120b (high), Terminus-2OpenAI53Complete19 of 19 tasks
gpt-oss-20b (high), Terminus-2OpenAI37Complete19 of 19 tasks
Kimi K2.6 (Reasoning), Terminus-2Moonshot AI79Complete19 of 19 tasks
Kimi K2.7 Code (Reasoning), Terminus-2Moonshot AI89Complete19 of 19 tasks
Kimi K3 (Reasoning), Terminus-2Moonshot AI89Complete19 of 19 tasks
Llama 3.3 70B (Non-reasoning), Terminus-2Meta0Complete19 of 19 tasks
Qwen 3.8 27B (xhigh), Terminus-2Alibaba (Qwen)100Still to run: 14 of 19 tasks5 of 19 tasks
Claude Haiku 4.5 (Reasoning), Claude CodeAnthropic74Complete19 of 19 tasks
GLM-5.3 (max), Claude CodeZ.ai95Complete19 of 19 tasks
Claude Opus 5.5 (medium), Claude CodeAnthropic95Complete19 of 19 tasks
Claude Opus 5.5 (high), Claude CodeAnthropic100Complete19 of 19 tasks
Claude Sonnet 5.5 (medium), Claude CodeAnthropic95Complete19 of 19 tasks
Claude Sonnet 5.5 (high), Claude CodeAnthropic100Complete19 of 19 tasks
Claude Haiku 5.5 (medium), Claude CodeAnthropic95Complete19 of 19 tasks
Claude Haiku 5.5 (high), Claude CodeAnthropic84Complete19 of 19 tasks
Claude Opus 5.5 (xhigh), Claude CodeAnthropic100Complete19 of 19 tasks
Claude Sonnet 5.5 (xhigh), Claude CodeAnthropic95Complete19 of 19 tasks
Claude Haiku 5.5 (xhigh), Claude CodeAnthropic95Complete19 of 19 tasks
Claude Opus 5 (high), Claude CodeAnthropic89Complete19 of 19 tasks
Claude Sonnet 5 (high), Claude CodeAnthropic95Complete19 of 19 tasks
Claude Opus 4.8 (high), Claude CodeAnthropic100Complete19 of 19 tasks
Claude Opus 4.7 (high), Claude CodeAnthropic100Complete19 of 19 tasks
GPT-6.1 Sol (medium), CodexOpenAI100Complete19 of 19 tasks
GPT-6.1 Sol (high), CodexOpenAI100Complete19 of 19 tasks
GPT-6.1 Sol (xhigh), CodexOpenAI95Complete19 of 19 tasks
GPT-6 Luna (xhigh), CodexOpenAI100Complete19 of 19 tasks
GPT-6 Sol (high), CodexOpenAI100Complete19 of 19 tasks
GPT-6 Luna (medium), CodexOpenAI89Complete19 of 19 tasks
GPT-6 Luna (high), CodexOpenAI100Complete19 of 19 tasks
GPT-5.6 Sol (high), CodexOpenAI89Complete19 of 19 tasks
GPT-5.6 Terra (high), CodexOpenAI95Complete19 of 19 tasks
GPT-5.6 Luna (medium), CodexOpenAI100Complete19 of 19 tasks
GPT-5.6 Luna (high), CodexOpenAI100Complete19 of 19 tasks
GPT-5.6 Luna (xhigh), CodexOpenAI100Still to run: 4 of 19 tasks15 of 19 tasks
GPT-6 Luna (low), CodexOpenAI79Complete19 of 19 tasks
GPT-5.6 Luna (low), CodexOpenAI95Complete19 of 19 tasks

Guided workflows includes GeoAgentBench and GeoAnalystBench, a group of the Geospatial Intelligence Index, not part of the Geospatial Agent Index. See the methodology for why.