Multimodal AI Models
29 LLMs that can process more than just text. Listed by declared capability breadth, without an inferred model-level quality rank.
0
Vision (Image Input)
3
Audio Input
0
Video Input
19
Text Only
Multimodal Models
| # | Model | Modalities | Quality evidence |
|---|---|---|---|
| 1 | Gemini 2.5 Flash (batch) Google | Text vision Audio | Not published |
| 2 | Gemini 2.5 Flash Lite (batch) Google | Text vision Audio | Not published |
| 3 | Gemini 2.5 Pro Google | Text vision Audio | Not published |
| 4 | Claude Haiku 4.5 (batch) Anthropic | Text vision | Not published |
| 5 | Claude Opus 4.5 (batch) Anthropic | Text vision | Not published |
| 6 | Claude Opus 4.6 (batch) Anthropic | Text vision | Not published |
| 7 | Claude Sonnet 4 Anthropic | Text vision | Not published |
| 8 | Claude Sonnet 4.5 (batch) Anthropic | Text vision | Not published |
| 9 | Claude Sonnet 4.6 (batch) Anthropic | Text vision | Not published |
| 10 | GPT-4.1 (batch) OpenAI | Text vision | Not published |
| 11 | GPT-4.1 Mini (batch) OpenAI | Text vision | Not published |
| 12 | GPT-4.1 Nano (batch) OpenAI | Text vision | Not published |
| 13 | GPT-4o (batch) OpenAI | Text vision | Not published |
| 14 | GPT-4o-mini (batch) OpenAI | Text vision | Not published |
| 15 | GPT-5 (batch) OpenAI | Text vision | Not published |
| 16 | GPT-5 Nano (batch) OpenAI | Text vision | Not published |
| 17 | GPT-5 Pro (batch) OpenAI | Text vision | Not published |
| 18 | GPT-5.2 OpenAI | Text vision | Not published |
| 19 | GPT-5.2 Pro (batch) OpenAI | Text vision | Not published |
| 20 | Llama 4 Maverick OSS Meta | Text vision | Not published |
| 21 | Llama 4 Scout OSS Meta | Text vision | Not published |
| 22 | Mistral Small 3.1 24B OSS Mistral | Text vision | Not published |
| 23 | Nova Lite 1.0 Amazon | Text vision | Not published |
| 24 | Nova Pro 1.0 Amazon | Text vision | Not published |
| 25 | o1 OpenAI | Text vision | Not published |
| 26 | O3 OpenAI | Text vision | Not published |
| 27 | O3 Pro OpenAI | Text vision | Not published |
| 28 | O4 Mini OpenAI | Text vision | Not published |
| 29 | Sonar Perplexity | Text vision | Not published |
Text-Only Models
These models process text input only. Listed by release date; no model-level quality score is inferred.
| # | Model | Quality evidence |
|---|---|---|
| 1 | DeepSeek V3.2 OSS DeepSeek | Not published |
| 2 | Mistral Large OSS Mistral | Not published |
| 3 | Reka Flash 3 Reka | Not published |
| 4 | R1 0528 OSS DeepSeek | Not published |
| 5 | Command A OSS Cohere | Not published |
| 6 | o3 Mini (batch) OpenAI | Not published |
| 7 | DeepSeek R1 OSS DeepSeek | Not published |
| 8 | DeepSeek V3 OSS DeepSeek | Not published |
| 9 | Phi 4 OSS Microsoft | Not published |
| 10 | Llama 3.3 70B Instruct OSS Meta | Not published |
| 11 | Nova Micro 1.0 Amazon | Not published |
| 12 | Qwen2.5 Coder 32B Instruct OSS Alibaba | Not published |
| 13 | Qwen2.5 72B Instruct OSS Alibaba | Not published |
| 14 | Command R+ (08-2024) OSS Cohere | Not published |
| 15 | Llama 3.1 70B Instruct OSS Meta | Not published |
| 16 | Llama 3.1 8B Instruct OSS Meta | Not published |
| 17 | Mistral Nemo OSS Mistral | Not published |
| 18 | Command R7B (12-2024) OSS Cohere | Not published |
| 19 | Command R (08-2024) OSS Cohere | Not published |
What are multimodal AI models?
Multimodal models can process multiple types of input — not just text, but also images, audio, and video. This enables use cases like:
- Vision: Analyse images, read charts, describe photos, OCR documents
- Audio: Transcribe speech, understand tone, process music
- Video: Summarise videos, extract key frames, answer questions about video content