Axis Spatial

Which model for which geospatial job

The best AI model for each kind of geospatial job, at what cost: the best entry overall, the best Claude Sonnet entry, the best under $0.02 a task and the best open-weights model. Data as of 2026-10-26.

Picks

Score: share of the evaluation’s tasks passed, one attempt per task. Cost is the entry’s average per task in that evaluation: the tokens it used at the model developer’s public list price, in that harness at that reasoning level. Your cost will differ with your prompts and caching. Reading images without tools is one model call per question, so its cost is per question. Sonnet-class: the best Claude Sonnet entry, the tier most teams start from. Only entries that have attempted every task of the evaluation are picked.

Find a model

Choose a job, a budget and a harness; the first entry left for each job is the best match.

Every entry with a complete run, by job
EntryJobHarnessScoreCost per taskOpen weights
Claude Opus 5.5 (high), Claude CodeComputing over satellite productsClaude Code79$0.18No
Claude Opus 5.5 (xhigh), Claude CodeComputing over satellite productsClaude Code77$0.40No
GPT-6 Sol (high), CodexComputing over satellite productsCodex75$0.10No
GLM-5.3 (max), Claude CodeComputing over satellite productsClaude Code74$0.98Yes
Claude Opus 5 (high), Claude CodeComputing over satellite productsClaude Code74$0.48No
Claude Haiku 5.5 (high), Claude CodeComputing over satellite productsClaude Code72$0.010No
GPT-6.1 Sol (medium), CodexComputing over satellite productsCodex72$0.056No
GPT-6.1 Sol (high), CodexComputing over satellite productsCodex72$0.072No
GPT-6.1 Sol (xhigh), CodexComputing over satellite productsCodex72$0.11No
DeepSeek V4 Flash (high), Terminus-2Computing over satellite productsTerminus-272$0.13Yes
Claude Opus 5.5 (medium), Claude CodeComputing over satellite productsClaude Code72$0.14No
Kimi K3 (Reasoning), Terminus-2Computing over satellite productsTerminus-272$0.41Yes
Claude Sonnet 5.5 (high), Claude CodeComputing over satellite productsClaude Code71$0.077No
Kimi K2.7 Code (Reasoning), Kimi CodeComputing over satellite productsKimi Code71$0.38Yes
Claude Opus 4.8 (high), Claude CodeComputing over satellite productsClaude Code71$0.44No
Kimi K2.7 Code (Reasoning), Terminus-2Computing over satellite productsTerminus-269$0.25Yes
Claude Opus 4.7 (high), Claude CodeComputing over satellite productsClaude Code69$0.58No
Claude Haiku 5.5 (xhigh), Claude CodeComputing over satellite productsClaude Code68$0.021No
Nemotron 3 Ultra (high), Terminus-2Computing over satellite productsTerminus-268$0.11Yes
Claude Sonnet 5.5 (xhigh), Claude CodeComputing over satellite productsClaude Code68$0.16No
GPT-5.6 Sol (high), CodexComputing over satellite productsCodex68$0.26No
Claude Sonnet 5 (high), Claude CodeComputing over satellite productsClaude Code68$0.32No
DeepSeek V4 Pro (high), Terminus-2Computing over satellite productsTerminus-268$0.57Yes
Claude Haiku 5.5 (medium), Claude CodeComputing over satellite productsClaude Code68$0.0070No
Kimi K2.6 (Reasoning), Terminus-2Computing over satellite productsTerminus-268$0.21Yes
Qwen 3.8 27B (xhigh), Qwen CodeComputing over satellite productsQwen Code68$0.32Yes
GLM-5.2 (high), Terminus-2Computing over satellite productsTerminus-268$0.39Yes
GLM-5.2 (max), Terminus-2Computing over satellite productsTerminus-268$0.48Yes
GPT-6 Luna (xhigh), CodexComputing over satellite productsCodex67$0.0087No
GPT-5.6 Luna (high), CodexComputing over satellite productsCodex67$0.023No
DeepSeek V4 Pro (max), Terminus-2Computing over satellite productsTerminus-267$0.76Yes
DeepSeek V4 Flash (max), Terminus-2Computing over satellite productsTerminus-266$0.11Yes
GPT-5.6 Luna (xhigh), CodexComputing over satellite productsCodex65$0.033No
GLM-5.3 (max), Terminus-2Computing over satellite productsTerminus-264$0.83Yes
GPT-6 Luna (high), CodexComputing over satellite productsCodex63$0.0046No
Kimi K2.6 (Reasoning), Kimi CodeComputing over satellite productsKimi Code63$0.60Yes
GPT-5.6 Terra (high), CodexComputing over satellite productsCodex62$0.12No
Claude Sonnet 5.5 (medium), Claude CodeComputing over satellite productsClaude Code58$0.054No
GLM-5.2 (Non-reasoning), Terminus-2Computing over satellite productsTerminus-258$0.15Yes
GLM-5.3 (low), Terminus-2Computing over satellite productsTerminus-257$0.62Yes
GPT-5.6 Luna (medium), CodexComputing over satellite productsCodex56$0.010No
Gemini 3.8 Flash (low), Terminus-2Computing over satellite productsTerminus-254$0.20No
Gemini 3.1 Flash-Lite (high), Terminus-2Computing over satellite productsTerminus-253$0.043No
GLM-5.3-Flash (low), Terminus-2Computing over satellite productsTerminus-253$0.13Yes
Claude Haiku 4.5 (Reasoning), Claude CodeComputing over satellite productsClaude Code51$0.14No
GLM-5.3-Flash (max), Terminus-2Computing over satellite productsTerminus-251$0.23Yes
gpt-oss-120b (high), Terminus-2Computing over satellite productsTerminus-250$0.035Yes
GPT-6 Luna (medium), CodexComputing over satellite productsCodex50$0.0028No
gpt-oss-120b (high), CodexComputing over satellite productsCodex49$0.65Yes
Nemotron 3.5 Super VL 120B (Reasoning), Terminus-2Computing over satellite productsTerminus-246$0.0099Yes
Gemma 4 26B A4B (Reasoning), Terminus-2Computing over satellite productsTerminus-246$0.027Yes
GPT-5.6 Luna (low), CodexComputing over satellite productsCodex44$0.0062No
GLM-4.7-Flash (Reasoning), Terminus-2Computing over satellite productsTerminus-244$0.047Yes
GPT-6 Luna (low), CodexComputing over satellite productsCodex44$0.0040No
Qwen 3.8 27B (xhigh), Terminus-2Computing over satellite productsTerminus-244$0.11Yes
Gemini 3.1 Flash-Lite (low), Terminus-2Computing over satellite productsTerminus-243$0.023No
MiniMax M3.1 Flash (max), Terminus-2Computing over satellite productsTerminus-239$0.019Yes
GLM-4.7-Flash (Reasoning), Claude CodeComputing over satellite productsClaude Code38$0.87Yes
Gemini 3.1 Flash-Lite (minimal), Terminus-2Computing over satellite productsTerminus-234$0.0082No
gpt-oss-20b (high), Terminus-2Computing over satellite productsTerminus-232$0.020Yes
Mistral Large 3 (Non-reasoning), Terminus-2Computing over satellite productsTerminus-231$0.047Yes
Nemotron 3 Nano Omni 30B (Reasoning), Terminus-2Computing over satellite productsTerminus-229$0.075Yes
gpt-oss-120b (low), Terminus-2Computing over satellite productsTerminus-228$0.0054Yes
Llama 3.3 70B (Non-reasoning), Terminus-2Computing over satellite productsTerminus-29$0.021Yes
gpt-oss-20b (high), CodexComputing over satellite productsCodex6$0.10Yes
Claude Opus 5.5 (xhigh), Claude CodeMaps and country or city dataClaude Code83$0.32No
Claude Sonnet 5.5 (high), Claude CodeMaps and country or city dataClaude Code81$0.062No
Claude Sonnet 5.5 (xhigh), Claude CodeMaps and country or city dataClaude Code81$0.15No
GPT-6.1 Sol (medium), CodexMaps and country or city dataCodex79$0.058No
GPT-6.1 Sol (xhigh), CodexMaps and country or city dataCodex79$0.13No
Claude Opus 5.5 (medium), Claude CodeMaps and country or city dataClaude Code79$0.14No
Claude Opus 5.5 (high), Claude CodeMaps and country or city dataClaude Code79$0.17No
Claude Haiku 5.5 (xhigh), Claude CodeMaps and country or city dataClaude Code77$0.026No
Claude Sonnet 5 (high), Claude CodeMaps and country or city dataClaude Code77$0.24No
GPT-6.1 Sol (high), CodexMaps and country or city dataCodex77$0.081No
Claude Sonnet 5.5 (medium), Claude CodeMaps and country or city dataClaude Code76$0.057No
GPT-6 Sol (high), CodexMaps and country or city dataCodex75$0.12No
Claude Opus 5 (high), Claude CodeMaps and country or city dataClaude Code75$0.40No
GPT-6 Luna (high), CodexMaps and country or city dataCodex74$0.0054No
Gemini 3.1 Pro (preview) (low), Terminus-2Maps and country or city dataTerminus-274$0.055No
Gemini 3.8 Flash (low), Terminus-2Maps and country or city dataTerminus-274$0.12No
GPT-6 Luna (xhigh), CodexMaps and country or city dataCodex73$0.010No
Gemini 3.1 Pro (preview) (high), Terminus-2Maps and country or city dataTerminus-273$0.16No
Claude Opus 4.8 (high), Claude CodeMaps and country or city dataClaude Code73$0.29No
Claude Opus 4.7 (high), Claude CodeMaps and country or city dataClaude Code73$0.34No
Claude Haiku 5.5 (high), Claude CodeMaps and country or city dataClaude Code72$0.011No
Claude Haiku 5.5 (medium), Claude CodeMaps and country or city dataClaude Code71$0.0082No
GPT-5.6 Sol (high), CodexMaps and country or city dataCodex70$0.32No
GPT-5.6 Luna (xhigh), CodexMaps and country or city dataCodex68$0.027No
GPT-5.6 Luna (high), CodexMaps and country or city dataCodex67$0.022No
GLM-5.2 (max), Terminus-2Maps and country or city dataTerminus-267$0.41Yes
GPT-5.6 Terra (high), CodexMaps and country or city dataCodex67$0.13No
GLM-5.2 (high), Terminus-2Maps and country or city dataTerminus-267$0.22Yes
DeepSeek V4 Pro (max), Terminus-2Maps and country or city dataTerminus-267$0.32Yes
Qwen 3.8 27B (xhigh), Terminus-2Maps and country or city dataTerminus-266$0.14Yes
GLM-5.2 (max), Claude CodeMaps and country or city dataClaude Code66$0.40Yes
DeepSeek V4 Pro (high), Terminus-2Maps and country or city dataTerminus-265$0.30Yes
DeepSeek V4 Flash (high), Terminus-2Maps and country or city dataTerminus-265$0.061Yes
Nemotron 3 Ultra (high), Terminus-2Maps and country or city dataTerminus-265$0.065Yes
GLM-5.3 (max), Terminus-2Maps and country or city dataTerminus-265$0.37Yes
GLM-5.3 (max), Claude CodeMaps and country or city dataClaude Code65$0.45Yes
Kimi K2.6 (Reasoning), Terminus-2Maps and country or city dataTerminus-264$0.065Yes
Kimi K2.7 Code (Reasoning), Kimi CodeMaps and country or city dataKimi Code64$0.19Yes
Kimi K2.6 (Reasoning), Kimi CodeMaps and country or city dataKimi Code64$0.19Yes
Kimi K3 (Reasoning), Terminus-2Maps and country or city dataTerminus-264$0.21Yes
Gemma 4 26B A4B (Reasoning), Terminus-2Maps and country or city dataTerminus-263$0.015Yes
GPT-5.6 Luna (medium), CodexMaps and country or city dataCodex63$0.012No
GPT-5.6 Luna (low), CodexMaps and country or city dataCodex62$0.0067No
DeepSeek V4 Flash (max), Terminus-2Maps and country or city dataTerminus-262$0.074Yes
Kimi K2.7 Code (Reasoning), Terminus-2Maps and country or city dataTerminus-262$0.11Yes
Qwen 3.8 27B (xhigh), Qwen CodeMaps and country or city dataQwen Code62$0.22Yes
Gemini 3.1 Flash-Lite (high), Terminus-2Maps and country or city dataTerminus-261$0.016No
GLM-5.2 (Non-reasoning), Terminus-2Maps and country or city dataTerminus-260$0.12Yes
GLM-5.3 (low), Terminus-2Maps and country or city dataTerminus-260$0.22Yes
GPT-6 Luna (medium), CodexMaps and country or city dataCodex57$0.0030No
GLM-5.3-Flash (low), Terminus-2Maps and country or city dataTerminus-257$0.027Yes
GLM-5.3-Flash (max), Terminus-2Maps and country or city dataTerminus-257$0.079Yes
Claude Sonnet 4.6 (high), Terminus-2Maps and country or city dataTerminus-257$0.20No
gpt-oss-120b (high), CodexMaps and country or city dataCodex55$0.16Yes
Nemotron 3.5 Super VL 120B (Reasoning), Terminus-2Maps and country or city dataTerminus-253$0.011Yes
Claude Haiku 4.5 (Reasoning), Claude CodeMaps and country or city dataClaude Code53$0.095No
GPT-6 Luna (low), CodexMaps and country or city dataCodex51$0.0035No
Gemini 3.1 Flash-Lite (low), Terminus-2Maps and country or city dataTerminus-251$0.010No
Gemini 3.8 Flash (high), Terminus-2Maps and country or city dataTerminus-251$0.31No
gpt-oss-120b (high), Terminus-2Maps and country or city dataTerminus-245$0.020Yes
Gemini 3.1 Flash-Lite (minimal), Terminus-2Maps and country or city dataTerminus-244$0.0055No
GLM-4.7-Flash (Reasoning), Terminus-2Maps and country or city dataTerminus-242$0.029Yes
GLM-4.7-Flash (Reasoning), Claude CodeMaps and country or city dataClaude Code39$0.082Yes
gpt-oss-20b (high), Terminus-2Maps and country or city dataTerminus-238$0.018Yes
gpt-oss-120b (low), Terminus-2Maps and country or city dataTerminus-237$0.0065Yes
Mistral Large 3 (Non-reasoning), Terminus-2Maps and country or city dataTerminus-229$0.024Yes
Nemotron 3 Nano Omni 30B (Reasoning), Terminus-2Maps and country or city dataTerminus-221$0.15Yes
gpt-oss-20b (high), CodexMaps and country or city dataCodex10$0.12Yes
Llama 3.3 70B (Non-reasoning), Terminus-2Maps and country or city dataTerminus-26$0.029Yes
GPT-6.1 Sol (xhigh), CodexSpatial database queriesCodex94$0.044No
Claude Opus 5.5 (xhigh), Claude CodeSpatial database queriesClaude Code94$0.11No
GPT-6.1 Sol (medium), CodexSpatial database queriesCodex93$0.033No
GPT-6.1 Sol (high), CodexSpatial database queriesCodex93$0.037No
Claude Opus 5 (high), Claude CodeSpatial database queriesClaude Code91$0.18No
Claude Opus 5.5 (medium), Claude CodeSpatial database queriesClaude Code90$0.075No
Claude Opus 5.5 (high), Claude CodeSpatial database queriesClaude Code90$0.083No
GPT-6 Sol (high), CodexSpatial database queriesCodex89$0.046No
Claude Sonnet 5.5 (xhigh), Claude CodeSpatial database queriesClaude Code87$0.057No
GPT-5.6 Sol (high), CodexSpatial database queriesCodex87$0.11No
Gemini 3.8 Flash (low), Terminus-2Spatial database queriesTerminus-286$0.071No
GPT-5.6 Luna (xhigh), CodexSpatial database queriesCodex85$0.010No
Claude Opus 4.8 (high), Claude CodeSpatial database queriesClaude Code84$0.15No
Claude Haiku 5.5 (high), Claude CodeSpatial database queriesClaude Code83$0.0058No
Claude Sonnet 5.5 (high), Claude CodeSpatial database queriesClaude Code83$0.039No
GLM-5.2 (max), Claude CodeSpatial database queriesClaude Code83$0.20Yes
GLM-5.3 (max), Terminus-2Spatial database queriesTerminus-283$0.15Yes
Kimi K2.7 Code (Reasoning), Terminus-2Spatial database queriesTerminus-282$0.045Yes
Claude Haiku 5.5 (xhigh), Claude CodeSpatial database queriesClaude Code81$0.0078No
GLM-5.3 (max), Claude CodeSpatial database queriesClaude Code81$0.23Yes
GPT-5.6 Luna (high), CodexSpatial database queriesCodex81$0.0093No
Kimi K2.7 Code (Reasoning), Kimi CodeSpatial database queriesKimi Code80$0.069Yes
GLM-5.2 (max), Terminus-2Spatial database queriesTerminus-280$0.18Yes
Kimi K3 (Reasoning), Terminus-2Spatial database queriesTerminus-280$0.15Yes
GPT-5.6 Luna (medium), CodexSpatial database queriesCodex80$0.0071No
Kimi K2.6 (Reasoning), Kimi CodeSpatial database queriesKimi Code80$0.14Yes
GPT-5.6 Terra (high), CodexSpatial database queriesCodex80$0.069No
GLM-5.3 (low), Terminus-2Spatial database queriesTerminus-280$0.11Yes
DeepSeek V4 Pro (max), Terminus-2Spatial database queriesTerminus-280$0.15Yes
GLM-5.3-Flash (max), Terminus-2Spatial database queriesTerminus-280$0.030Yes
DeepSeek V4 Flash (max), Terminus-2Spatial database queriesTerminus-280$0.029Yes
DeepSeek V4 Flash (high), Terminus-2Spatial database queriesTerminus-279$0.032Yes
Claude Haiku 5.5 (medium), Claude CodeSpatial database queriesClaude Code79$0.0046No
Claude Opus 4.7 (high), Claude CodeSpatial database queriesClaude Code79$0.21No
Claude Sonnet 5.5 (medium), Claude CodeSpatial database queriesClaude Code79$0.035No
GLM-5.2 (high), Terminus-2Spatial database queriesTerminus-278$0.13Yes
GPT-6 Luna (xhigh), CodexSpatial database queriesCodex78$0.0040No
Claude Sonnet 5 (high), Claude CodeSpatial database queriesClaude Code77$0.10No
DeepSeek V4 Pro (high), Terminus-2Spatial database queriesTerminus-277$0.15Yes
Kimi K2.6 (Reasoning), Terminus-2Spatial database queriesTerminus-277$0.044Yes
GPT-6 Luna (high), CodexSpatial database queriesCodex77$0.0028No
GLM-5.3-Flash (max), Claude CodeSpatial database queriesClaude Code76$0.046Yes
GPT-5.6 Luna (low), CodexSpatial database queriesCodex76$0.0050No
GLM-5.3-Flash (low), Terminus-2Spatial database queriesTerminus-276$0.016Yes
Qwen 3.8 27B (xhigh), Qwen CodeSpatial database queriesQwen Code75$0.16Yes
DeepSeek V4 Flash (high), CodexSpatial database queriesCodex74$0.043Yes
Qwen 3.8 27B (xhigh), Terminus-2Spatial database queriesTerminus-274$0.074Yes
GLM-5.2 (Non-reasoning), Terminus-2Spatial database queriesTerminus-272$0.059Yes
Nemotron 3 Ultra (high), Terminus-2Spatial database queriesTerminus-270$0.049Yes
Claude Haiku 4.5 (Reasoning), Claude CodeSpatial database queriesClaude Code69$0.067No
GPT-6 Luna (medium), CodexSpatial database queriesCodex65$0.0023No
GPT-6 Luna (low), CodexSpatial database queriesCodex59$0.0025No
Nemotron 3.5 Super VL 120B (Reasoning), Terminus-2Spatial database queriesTerminus-259$0.0086Yes
Gemma 4 26B A4B (Reasoning), Terminus-2Spatial database queriesTerminus-258$0.022Yes
Gemini 3.1 Flash-Lite (high), Terminus-2Spatial database queriesTerminus-255$0.035No
Gemini 3.1 Flash-Lite (low), Terminus-2Spatial database queriesTerminus-249$0.021No
GLM-4.7-Flash (Reasoning), Claude CodeSpatial database queriesClaude Code48$0.074Yes
GLM-4.7-Flash (Reasoning), Terminus-2Spatial database queriesTerminus-243$0.036Yes
gpt-oss-120b (high), Terminus-2Spatial database queriesTerminus-242$0.025Yes
Mistral Large 3 (Non-reasoning), Terminus-2Spatial database queriesTerminus-239$0.016Yes
gpt-oss-20b (high), Terminus-2Spatial database queriesTerminus-230$0.016Yes
gpt-oss-120b (low), Terminus-2Spatial database queriesTerminus-229$0.0030Yes
gpt-oss-120b (high), CodexSpatial database queriesCodex25$0.12Yes
Gemini 3.1 Flash-Lite (minimal), Terminus-2Spatial database queriesTerminus-218$0.0063No
gpt-oss-20b (high), CodexSpatial database queriesCodex12$0.068Yes
Llama 3.3 70B (Non-reasoning), Terminus-2Spatial database queriesTerminus-210$0.020Yes
Nemotron 3 Nano Omni 30B (Reasoning), Terminus-2Spatial database queriesTerminus-29$0.15Yes
Claude Opus 5.5 (xhigh), Claude CodeMeasuring objects in imagesClaude Code97$0.17No
Claude Sonnet 5.5 (xhigh), Claude CodeMeasuring objects in imagesClaude Code89$0.068No
GPT-6.1 Sol (high), CodexMeasuring objects in imagesCodex86$0.034No
Claude Opus 5 (high), Claude CodeMeasuring objects in imagesClaude Code86$0.65No
Claude Sonnet 5 (high), Claude CodeMeasuring objects in imagesClaude Code81$0.48No
Claude Haiku 5.5 (xhigh), Claude CodeMeasuring objects in imagesClaude Code80$0.013No
GPT-6 Sol (high), CodexMeasuring objects in imagesCodex80$0.053No
GLM-5.3-Flash (max), Claude CodeMeasuring objects in imagesClaude Code78$0.31Yes
Claude Opus 4.7 (high), Claude CodeMeasuring objects in imagesClaude Code77$1.04No
GPT-5.6 Sol (high), CodexMeasuring objects in imagesCodex75$0.19No
Claude Opus 4.8 (high), Claude CodeMeasuring objects in imagesClaude Code73$0.51No
GPT-5.6 Luna (xhigh), CodexMeasuring objects in imagesCodex72$0.032No
GPT-6 Luna (xhigh), CodexMeasuring objects in imagesCodex70$0.0052No
GPT-5.6 Luna (medium), CodexMeasuring objects in imagesCodex69$0.0083No
GPT-5.6 Luna (high), CodexMeasuring objects in imagesCodex69$0.019No
GPT-5.6 Terra (high), CodexMeasuring objects in imagesCodex69$0.11No
GPT-6 Luna (high), CodexMeasuring objects in imagesCodex67$0.0030No
GPT-6 Luna (medium), CodexMeasuring objects in imagesCodex64$0.0019No
GPT-5.6 Luna (low), CodexMeasuring objects in imagesCodex64$0.0043No
GPT-6 Luna (low), CodexMeasuring objects in imagesCodex28$0.0017No
Claude Haiku 4.5 (Reasoning), Claude CodeMeasuring objects in imagesClaude Code19$0.13No
Claude Opus 5.5 (high)Reading images without toolsNo tools (one call)59No dataNo
Claude Sonnet 5.5 (high)Reading images without toolsNo tools (one call)59No dataNo
Claude Opus 5.5 (xhigh)Reading images without toolsNo tools (one call)59No dataNo
GPT-5.6 Sol (high)Reading images without toolsNo tools (one call)58No dataNo
Claude Sonnet 5.5 (xhigh)Reading images without toolsNo tools (one call)57No dataNo
Claude Opus 5.5 (medium)Reading images without toolsNo tools (one call)57No dataNo
GPT-6.1 Sol (xhigh)Reading images without toolsNo tools (one call)56No dataNo
GPT-6 Sol (high)Reading images without toolsNo tools (one call)56No dataNo
Claude Sonnet 5.5 (medium)Reading images without toolsNo tools (one call)56No dataNo
GPT-6.1 Sol (medium)Reading images without toolsNo tools (one call)55No dataNo
GPT-5.6 Terra (high)Reading images without toolsNo tools (one call)54No dataNo
Claude Opus 5 (high)Reading images without toolsNo tools (one call)53No dataNo
Qwen 3.8 27B (xhigh)Reading images without toolsNo tools (one call)53No dataYes
GPT-6.1 Sol (high)Reading images without toolsNo tools (one call)53No dataNo
GPT-6 Luna (high)Reading images without toolsNo tools (one call)53No dataNo
Claude Haiku 5.5 (xhigh)Reading images without toolsNo tools (one call)52No dataNo
GPT-5.6 Luna (xhigh)Reading images without toolsNo tools (one call)52No dataNo
GPT-5.6 Luna (high)Reading images without toolsNo tools (one call)51No dataNo
GPT-6 Luna (xhigh)Reading images without toolsNo tools (one call)50No dataNo
GPT-5.6 Luna (low)Reading images without toolsNo tools (one call)50No dataNo
Claude Opus 4.8 (high)Reading images without toolsNo tools (one call)50No dataNo
Gemini 3.8 Flash (low)Reading images without toolsNo tools (one call)50No dataNo
GPT-6 Luna (medium)Reading images without toolsNo tools (one call)50No dataNo
Kimi K3 (Reasoning)Reading images without toolsNo tools (one call)48No dataYes
GPT-6 Luna (low)Reading images without toolsNo tools (one call)48No dataNo
Claude Haiku 5.5 (high)Reading images without toolsNo tools (one call)47No dataNo
GPT-5.6 Luna (medium)Reading images without toolsNo tools (one call)47No dataNo
Claude Opus 4.7 (high)Reading images without toolsNo tools (one call)47No dataNo
Claude Sonnet 5 (high)Reading images without toolsNo tools (one call)47No dataNo
GLM-5.3-Flash (low)Reading images without toolsNo tools (one call)46No dataYes
Claude Haiku 5.5 (medium)Reading images without toolsNo tools (one call)45No dataNo
Kimi K2.6 (Reasoning)Reading images without toolsNo tools (one call)44No dataYes
GLM-5.3-Flash (max)Reading images without toolsNo tools (one call)44No dataYes
Kimi K2.7 Code (Reasoning)Reading images without toolsNo tools (one call)44No dataYes
Gemma 4 26B A4B (Reasoning)Reading images without toolsNo tools (one call)43No dataYes
Gemini 3.1 Flash-Lite (minimal)Reading images without toolsNo tools (one call)41No dataNo
Gemini 3.1 Flash-Lite (high)Reading images without toolsNo tools (one call)38No dataNo
Gemini 3.1 Flash-Lite (low)Reading images without toolsNo tools (one call)35No dataNo
Claude Haiku 4.5 (Reasoning)Reading images without toolsNo tools (one call)34No dataNo
Mistral Large 3 (Non-reasoning)Reading images without toolsNo tools (one call)33No dataYes
Mistral Small 3.1 24B (Non-reasoning)Reading images without toolsNo tools (one call)33No dataYes