What does a blank cell mean?

Nothing was listed. It does not mean no. The two are different answers and this table keeps them apart: a tick is a yes, a cross is a source that says no, and an empty cell is silence.

Why can I filter for models that lack tool calling, but not for models that lack web search?

Tool calling comes from a source that publishes both true and false, so a no is a measurement. The listed-capability columns come from a source that publishes only the capabilities a model has. It never writes a no, so this site cannot offer one.

Where do the modality columns come from?

From the input and output modality lists. A list that exists and omits image is a real no, because the source enumerated the modalities and image was not among them. A model with no list at all reads as not listed.

Is a capability the same as support at every host?

No. These are model-level facts as the sources describe them. A host can expose a subset. Open the model page to see the hosts.

Inputs, outputs and features

AI model capabilities

What each model takes in, what it puts out, and what the sources say it can do. A blank cell means nothing was listed. It never means no. All models.

2,540 models

✓ yes✕ no, and a source says so nothing listed

Model	Lab	Tools	Reasoning	Structured	Open weights	Image in
Gemini 3.5 Flash (Vertex AI)	Google	yes	yes	yes	no	yes
Gemini 3.8 Flash (Vertex AI)	Google	yes	yes	yes	no	yes
Gemini 3.7 Flash (Vertex AI)	Google	yes	yes	yes	no	yes
Gemini 3.6 Flash (Vertex AI)	Google	yes	yes	yes	no	yes
Gemini 3 Flash Preview (Vertex AI)	Google	yes	yes	yes	no	yes
Gemini 3.5 Flash Lite (Vertex AI)	Google	yes	yes	yes	no	yes
Gemini 4 Argon (Vertex AI)	Google	yes	yes	yes	no	yes
Claude Haiku 4.5 (latest)	Anthropic	yes	yes	yes	no	yes
Gemini 2.5 Flash	Google	yes	yes	yes	no	yes
Gemini 3.1 Flash Lite (Vertex AI)	Google	yes	yes	yes	no	yes
Gemini 2.5 Flash-Lite	Google	yes	yes	yes	no	yes
Claude Sonnet 4.6	Anthropic	yes	yes	yes	no	yes
GPT-5.6 Sol	OpenAI	yes	yes	yes	no	yes
Claude Sonnet 4.5 (latest)	Anthropic	yes	yes	yes	no	yes
Gemini 3.1 Pro Preview (Vertex AI)	Google	yes	yes	yes	no	yes
GPT-6 Luna	OpenAI	yes	yes	yes	no	yes
GPT-6 Astra	OpenAI	yes	yes	yes	no	yes
GPT-6 Sol	OpenAI	yes	yes	yes	no	yes
GPT-6.1 Sol	OpenAI	yes	yes	yes	no	yes
Gemini 2.5 Pro	Google	yes	yes	yes	no	yes
GPT-6.1 Terra	OpenAI	yes	yes	yes	no	yes
DeepSeek V4 Flash 0731 (Nebius)	Alibaba	yes	yes	yes	yes	yes
GLM-5.3	Alibaba	yes	yes	yes	yes	yes
GLM-5.2	Alibaba	yes	yes	yes	yes	yes
GLM-5.3-Flash	Zhipu	yes	yes	yes	yes	yes
GPT-5.5	OpenAI	yes	yes	yes	no	yes
Claude Opus 4.8	Anthropic	yes	yes	yes	no	yes
GPT-5.6 Luna	OpenAI	yes	yes	yes	no	yes
Claude Fable 5	Anthropic	yes	yes	yes	no	yes
Claude Sonnet 5	Anthropic	yes	yes	yes	no	yes
GPT-5.4	OpenAI	yes	yes	yes	no	yes
Claude Opus 5	Anthropic	yes	yes	yes	no	yes
GPT-5.6 Terra	OpenAI	yes	yes	yes	no	yes
Claude Fable 5.1	Anthropic	yes	yes	yes	no	yes
Claude Opus 4.5 (latest)	Anthropic	yes	yes	yes	no	yes
Claude Opus 5.5	Anthropic	yes	yes	yes	no	yes
GPT-5.4 mini	OpenAI	yes	yes	yes	no	yes
Claude Haiku 5.5	Anthropic	yes	yes	yes	no	yes
Claude Sonnet 5.5	Anthropic	yes	yes	yes	no	yes
GPT-5	OpenAI	yes	yes	yes	no	yes
GPT-5 Mini	OpenAI	yes	yes	yes	no	yes
GPT-5.4 nano	OpenAI	yes	yes	yes	no	yes
GPT-5.2	OpenAI	yes	yes	yes	no	yes
GPT-5.3 Codex	OpenAI	yes	yes	yes	no	yes
GPT-5 Nano	OpenAI	yes	yes	yes	no	yes
GPT-5.1	OpenAI	yes	yes	yes	no	yes
GPT-4o (2024-08-06)	OpenAI	yes	no	yes	no	yes
GPT-5.4 Pro	OpenAI	yes	yes	yes	no	yes
GPT-5.2 Codex (Azure)	OpenAI	yes	yes	yes	no	yes
GPT-5.5 Pro	OpenAI	yes	yes	yes	no	yes
Muse Spark 1.2 (Meta)	Meta	yes	yes	yes	no	yes
GPT-5 Pro	OpenAI	yes	yes	yes	no	yes
Muse Spark 1.3 (Meta)	Meta	yes	yes	yes	no	yes
Muse Spark 1.1 (Meta)	Meta	yes	yes	yes	no	yes
DeepSeek V4 Pro 0813 (Nebius)	Alibaba	yes	yes	yes	yes	no
Kimi K3	Kimi	yes	yes	yes	yes	yes
DeepSeek V4.1 Flash (Nebius)	Alibaba	yes	yes	yes	yes	yes
Kimi K2.7 Code	Kimi	yes	yes	yes	yes	yes
GPT OSS 120B (IONOS)	OpenAI	yes	yes	yes	yes	yes
Claude Opus 4.7	Anthropic	yes	yes	yes	no	yes
GPT OSS 20B (Groq)	OpenAI	yes	yes	yes	yes	no
Claude Opus 4.6	Anthropic	yes	yes	yes	no	yes
Grok 4.6	xAI	yes	yes	yes	no	yes
DeepSeek R1 0528 (TensorX)	DeepSeek	yes	yes	yes	yes	no
Grok 4.7	xAI	yes	yes	yes	no	yes
Qwen3.6 35B A3B	Alibaba	yes	yes	yes	yes	yes
GPT-4.1	OpenAI	yes	no	yes	no	yes
Grok 4.5	xAI	yes	yes	yes	no	yes
GPT-4.1 mini	OpenAI	yes	no	yes	no	yes
DeepSeek-V3 (Deep Infra)	DeepSeek	yes	yes	yes	yes	no
GLM-4.5-Air	Zhipu	yes	yes	yes	yes	no
Kimi K2 Thinking (Amazon Bedrock)	Amazon	yes	yes	yes	yes	no
claude-opus-4-1-20250805	Anthropic	yes	yes	yes	no	yes
GPT-5.1 Codex (Azure)	OpenAI	yes	yes	yes	no	yes
DeepSeek V4 Flash Vision Exp	DeepSeek	yes	yes	yes	yes	yes
o4-mini	OpenAI	yes	yes	yes	no	yes
GPT-5.1 Codex Max (Azure)	OpenAI	yes	yes	yes	no	yes
Kimi K2.6	Kimi	yes	yes	yes	yes	yes
MiniMax-M3	MiniMax	yes	yes	yes	yes	yes
GLM-5.1	Zhipu	yes	yes	yes	yes	no
Qwen3.8 27B	Alibaba	yes	yes	yes	yes	yes
MiniMax-M2.5	MiniMax	yes	yes	yes	yes	no
Gemma 4 31B IT	Google	yes	yes	yes	yes	yes
GLM-4.7	Zhipu	yes	yes	yes	yes	no
GLM-4.6	Zhipu	yes	yes	yes	yes	no
Qwen3.6 27B	Alibaba	yes	yes	yes	yes	yes
MiMo V2.6 Flash	—	yes	yes	yes	yes	yes
MiMo V2.6 Pro	—	yes	yes	yes	yes	yes
GPT-4o mini	OpenAI	yes	no	yes	no	yes
MiniMax-M2.1	MiniMax	yes	yes	yes	yes	no
GLM-4.7-Flash (Deep Infra)	Zhipu	yes	yes	yes	yes	no
o3	OpenAI	yes	yes	yes	no	yes
DeepSeek V3.2	DeepSeek	yes	yes	yes	yes	yes
GPT-4.1 nano (Azure)	OpenAI	yes	no	yes	no	yes
GPT-5.1 Codex mini (Azure)	OpenAI	yes	yes	yes	no	yes
MiniMax-M2	MiniMax	yes	yes	yes	yes	no
Qwen3 Max (Deep Infra)	Alibaba	yes	yes	yes	no	no
DeepSeek V3 0324	DeepSeek	yes	no	yes	yes	no
MiniMax-M2.7	MiniMax	yes	yes	yes	yes	no
Qwen3.8 Max 0902	Alibaba	yes	yes	yes	yes	yes
Qwen3.8 Flash	Alibaba	yes	yes	yes	no	yes
Grok 4.3	xAI	yes	yes	yes	no	yes
Qwen3.7 Plus	Alibaba	yes	yes	yes	no	yes
Qwen3.5 397B-A17B	Alibaba	yes	yes	yes	yes	yes
Gemma 4 26B A4B IT	Google	yes	yes	yes	yes	yes
Mistral Large 3	Mistral	yes	yes	yes	yes	yes
Qwen3-Next 80B-A3B Instruct (Deep Infra)	Alibaba	yes	no	yes	yes	no
Hy3 (Deep Infra)	—	yes	yes	yes	yes	no
Mistral Small (latest)	Mistral	yes	yes	yes	yes	yes
Qwen3.8 2.4T A95B	Alibaba	yes	yes	yes	yes	yes
Inkling (Deep Infra)	Databricks	yes	yes	yes	yes	yes
GLM-4.5	Zhipu	yes	yes	yes	yes	no
Qwen3.5 122B-A10B	Alibaba	yes	yes	yes	yes	yes
Qwen3 VL 235B A22B Instruct (Deep Infra)	Alibaba	yes	no	yes	yes	yes
GLM-4.5V	Zhipu	yes	yes	yes	yes	yes
GLM-4.6V	Zhipu	yes	yes	yes	yes	yes
Muse Spark Open (Nebius)	Meta	yes	yes	yes	yes	yes
Kimi K2.7 Code Highspeed	Kimi	yes	yes	yes	yes	yes
Kimi K2.5 (TensorX)	Kimi	yes	yes	yes	yes	yes
GLM-5	Alibaba	yes	yes	yes	yes	no
Qwen3.7 Max	Alibaba	yes	yes	yes	no	no
Llama-3.3-70B-Instruct (IONOS)	Meta	yes	no	yes	yes	no
MiniMax-M3.1-Flash	MiniMax	yes	yes	yes	no	yes
MiMo-V2.5-Pro	—	yes	yes	yes	yes	no
Qwen3.5 35B-A3B	Alibaba	yes	yes	yes	yes	yes
Step 3.7 Flash (Deep Infra)	—	yes	yes	yes	yes	yes
Qwen3.5 9B	Alibaba	yes	yes	yes	yes	yes
Hy4 preview (Deep Infra)	—	yes	yes	yes	yes	no
Qwen3 235B-A22B Instruct 2507 (Nebius)	Alibaba	yes	no	yes	yes	no
Grok Build 0.1	xAI	yes	yes	yes	no	yes
GLM-5-Turbo	Zhipu	yes	yes	yes	no	no
MiniMax M2.7 Highspeed	MiniMax	yes	yes	yes	yes	no
GLM-5V-Turbo	Zhipu	yes	yes	yes	no	yes
Qwen3-Coder 30B-A3B Instruct (Scaleway)	Alibaba	yes	no	yes	yes	no
Qwen3-235B-A22B	Alibaba	yes	yes	yes	yes	no
Muse Glimmer 30B (Deep Infra)	Meta	yes	yes	yes	yes	yes
Nano Banana Pro (Vertex AI)	Google	yes	yes	yes	no	yes
Qwen/Qwen3-32B	Alibaba	yes	yes	yes	yes	no
Step 5 Preview (Deep Infra)	—	yes	yes	yes	yes	yes
Kimi K3.1 Pro	Kimi	yes	yes	yes	no	yes
Qwen3.5 27B	Alibaba	yes	yes	yes	yes	yes
Qwen3 Coder Next (IONOS)	Amazon	yes	yes	yes	yes	no
Codestral 25.08	Mistral	yes	no	yes	yes	no
Gemma 3 27B IT (Nebius)	Amazon	yes	yes	yes	yes	yes
Qwen3-30B-A3B	Alibaba	yes	yes	yes	yes	no
MiMo-V2.5 (Free)	—	yes	yes	yes	yes	yes
Qwen3.5 Plus	Alibaba	yes	yes	yes	no	yes
Qwen3.6 Plus	Alibaba	yes	yes	yes	no	yes
Qwen3.6 Flash	Alibaba	yes	yes	yes	no	yes
search	DeepSeek	not listed	not listed	not listed	not listed	not listed

What each column can say

The left group has a source that publishes both yes and no, so each column reports three states. Counted over 2,540 models:

Tools — 149 yes, 0 no, 1 not listed
Reasoning — 137 yes, 12 no, 1 not listed
Structured — 149 yes, 0 no, 1 not listed
Open weights — 70 yes, 79 no, 1 not listed
Image in — 116 yes, 33 no, 1 not listed
Audio in — 24 yes, 125 no, 1 not listed
Video in — 52 yes, 97 no, 1 not listed
PDF in — 97 yes, 52 no, 1 not listed
Image out — 7 yes, 142 no, 1 not listed
Audio out — 0 yes, 149 no, 1 not listed

The right group has no such source. Every value it carries is a yes — there is not one no in the corpus — because the source lists the capabilities a model has and stays silent about the rest. So those columns can be filtered for is listed and never for lacks. 148 of 2,540 models carry at least one listed capability.

Common questions

What does a blank cell mean?: Nothing was listed. It does not mean no. The two are different answers and this table keeps them apart: a tick is a yes, a cross is a source that says no, and an empty cell is silence.
Why can I filter for models that lack tool calling, but not for models that lack web search?: Tool calling comes from a source that publishes both true and false, so a no is a measurement. The listed-capability columns come from a source that publishes only the capabilities a model has. It never writes a no, so this site cannot offer one.
Where do the modality columns come from?: From the input and output modality lists. A list that exists and omits image is a real no, because the source enumerated the modalities and image was not among them. A model with no list at all reads as not listed.
Is a capability the same as support at every host?: No. These are model-level facts as the sources describe them. A host can expose a subset. Open the model page to see the hosts.