AI models: what each one is good for
363 models from 61 labs. Start with a lab or a use case, then open a model to see why you would pick it, what it costs and how much text fits in its context window.
Use them all in one app. Claude, ChatGPT, Gemini, Grok and DeepSeek with one account and one balance. Switch on any message.
Compare the modelsAll models in one appTools
By use case
- Frontier modelsThe highest-scoring models on the Artificial Analysis Intelligence Index.
- Models from Chinese labsDeepSeek, Qwen, Z.ai (GLM), Moonshot (Kimi), MiniMax and more.
- Long-context modelsContext windows of one million tokens or more: whole books, contracts or codebases in a single prompt.
- Models that understand imagesModels that take screenshots, photos, charts and scanned pages as input, for document understanding, OCR-style extraction and visual Q&A.
- Image generation modelsModels that create images from text prompts, for illustrations, covers and diagrams in documents.
- Speech and audio modelsSpeech-to-text, text-to-speech and audio-understanding models for transcription, voice agents and narration.
- Best-value modelsModels with a solid benchmark score at a low price per million tokens: the sensible default for high-volume drafting and summarising.
By lab
OpenAI58 modelsGPT and reasoning models, plus open-weight gpt-oss.
Qwen (Alibaba)40 modelsThe broadest open-weight family, from tiny to frontier.
Google36 modelsGemini: huge context windows and native multimodal input; Gemma for open weights.
Mistral AI20 modelsEuropean labs: efficient open and commercial models.
Z.ai (GLM)17 modelsGLM: strong agentic coding at low prices.
Anthropic16 modelsClaude: careful, strong at coding and long agentic work.
DeepSeek14 modelsVery low-cost open-weight reasoning and coding models.
NVIDIA11 modelsNemotron open models tuned for agents.
Meta8 modelsLlama: the most widely deployed open-weight family.
Moonshot AI (Kimi)8 modelsKimi: long-horizon agent and coding models.
xAI7 modelsGrok: real-time knowledge and large context.
Tencent7 modelsHunyuan models.
Cohere6 modelsEnterprise-focused models for retrieval and tools.
Xiaomi5 modelsMiMo reasoning models.
Perplexity5 modelsSearch-grounded answer models.
Amazon5 modelsNova models on AWS.
MiniMax4 modelsFast, low-cost agent models.
ByteDance Seed3 modelsSeed multimodal and reasoning models.
Models WordGPT runs today
Ling 3.0 Flash VL
Used in WordGPTinclusionai · Sep 10, 2026
TextImageVideoText131K context$0.060 in / $0.18 out per M
DeepSeek V4.1 Flash
Used in WordGPTDeepSeek · Sep 10, 2026
TextImageText1.05M context$0.14 in / $0.42 out per M
GLM 5.3 Flash
Used in WordGPTZ.ai (GLM) · Aug 26, 2026
TextImageVideoText1.05M contextScore 27.8 · $0.040 in / $0.078 out per M
GPT-5.6 Luna
Used in WordGPTOpenAI · Jul 9, 2026
Files / PDFImageTextText1.05M contextScore 23.8 · $0.2 in / $1.2 out per M
Gemini 3.1 Flash Lite
Used in WordGPTGoogle · May 7, 2026
TextImageVideoFiles / PDFAudioText1.05M context$0.25 in / $1.5 out per M
Newest
Qwen4 Max
Qwen (Alibaba) · Nov 16, 2026
TextImageVideoText1M context$1.5 in / $4.5 out per M
Ling 3.2 Flash (free)
inclusionai · Nov 13, 2026
TextImageText1.05M contextFree in / Free out per M
Gemini 4 Hydrogen
Google · Nov 12, 2026
AudioFiles / PDFImageTextVideoText1.05M contextScore 53.9 · $1.5 in / $7.5 out per M
ERNIE-5.5-Flash-Preview (free)
baidu · Nov 10, 2026
TextText1.02M contextFree in / Free out per M
ERNIE-Light-DT-1.0
baidu · Nov 10, 2026
TextText1.02M context$1.13 in / $3.94 out per M
Gemini 4 Carbon
Google · Nov 10, 2026
AudioFiles / PDFImageTextVideoText1.05M contextScore 60 · $2 in / $10 out per M
Try these models in WordGPT
WordGPT picks and switches models for you: drafting, editing and files in one place. Free plan, no credit card.
Try it free