Skip to content

Models Reference

This page lists models that expose structured output, reasoning, text verbosity, tokenization modality support, and built-in Anthropic web tools (web_search / web_fetch) where available. Model names prefer aliases where available; otherwise the table shows the minimal model string (including provider prefix when required). The provider column reflects the default provider used by fast-agent (models may become available with multiple providers in the future).

Model Provider Tokenizes Structured Output Reasoning Verbosity Built-in Web Tools
qwen-long aliyun Text json (schema) — — —
qwen-max aliyun Text, Vision json (object) — — —
qwen-plus aliyun Text, Vision json (object) — — —
qwen-turbo aliyun Text, Vision json (object) — — —
qwen3-max aliyun Text json (schema) — — —
claude-3-5-haiku-20241022 anthropic Text, Vision, Document tool_use — — web_search (web_search_20250305)
web_fetch (web_fetch_20250910)
claude-3-5-haiku-latest anthropic Text, Vision, Document tool_use — — web_search (web_search_20250305)
web_fetch (web_fetch_20250910)
claude-3-5-haiku anthropic Text, Vision, Document tool_use — — web_search (web_search_20250305)
web_fetch (web_fetch_20250910)
claude-fable-5-1 anthropic Text, Vision, Document json (schema) effort: auto, low, medium, high, xhigh, max
Example: claude-fable-5-1?reasoning=auto		 — web_search (web_search_20260209)
web_fetch (web_fetch_20260209)
beta: code-execution-web-tools-2026-02-09
claude-haiku-4-5-20251001 anthropic Text, Vision, Document json (schema) budget: low, medium, high, max, 0, 1024, 16000, 32000, off
Example: claude-haiku-4-5-20251001?reasoning=1024		 — web_search (web_search_20250305)
web_fetch (web_fetch_20250910)
claude-opus-4-0 anthropic Text, Vision, Document tool_use budget: low, medium, high, max, 0, 1024, 16000, 32000, off
Example: claude-opus-4-0?reasoning=1024		 — web_search (web_search_20250305)
web_fetch (web_fetch_20250910)
claude-opus-4-1 anthropic Text, Vision, Document json (schema) budget: low, medium, high, max, 0, 1024, 16000, 32000, off
Example: claude-opus-4-1?reasoning=1024		 — web_search (web_search_20250305)
web_fetch (web_fetch_20250910)
claude-opus-4-20250514 anthropic Text, Vision, Document tool_use budget: low, medium, high, max, 0, 1024, 16000, 32000, off
Example: claude-opus-4-20250514?reasoning=1024		 — web_search (web_search_20250305)
web_fetch (web_fetch_20250910)
claude-opus-4-5 anthropic Text, Vision, Document json (schema) budget: low, medium, high, max, 0, 1024, 16000, 32000, off
Example: claude-opus-4-5?reasoning=1024		 — web_search (web_search_20250305)
web_fetch (web_fetch_20250910)
claude-sonnet-4-0 anthropic Text, Vision, Document tool_use budget: low, medium, high, max, 0, 1024, 16000, 32000, off
Example: claude-sonnet-4-0?reasoning=1024		 — web_search (web_search_20250305)
web_fetch (web_fetch_20250910)
claude-sonnet-4-20250514 anthropic Text, Vision, Document tool_use budget: low, medium, high, max, 0, 1024, 16000, 32000, off
Example: claude-sonnet-4-20250514?reasoning=1024		 — web_search (web_search_20250305)
web_fetch (web_fetch_20250910)
claude-sonnet-4-5-20250929 anthropic Text, Vision, Document json (schema) budget: low, medium, high, max, 0, 1024, 16000, 32000, off
Example: claude-sonnet-4-5-20250929?reasoning=1024		 — web_search (web_search_20250305)
web_fetch (web_fetch_20250910)
claude-sonnet-4-5 anthropic Text, Vision, Document json (schema) budget: low, medium, high, max, 0, 1024, 16000, 32000, off
Example: claude-sonnet-4-5?reasoning=1024		 — web_search (web_search_20250305)
web_fetch (web_fetch_20250910)
claude anthropic Text, Vision, Document json (schema) effort: auto, low, medium, high, xhigh, max, off
Example: claude?reasoning=auto		 — web_search (web_search_20260209)
web_fetch (web_fetch_20260209)
beta: code-execution-web-tools-2026-02-09
fable anthropic Text, Vision, Document json (schema) effort: auto, low, medium, high, xhigh
Example: fable?reasoning=auto		 — web_search (web_search_20260209)
web_fetch (web_fetch_20260209)
beta: code-execution-web-tools-2026-02-09
haiku anthropic Text, Vision, Document json (schema) budget: low, medium, high, max, 0, 1024, 16000, 32000, off
Example: haiku?reasoning=1024		 — web_search (web_search_20250305)
web_fetch (web_fetch_20250910)
opus46 anthropic Text, Vision, Document json (schema) effort: auto, low, medium, high, max, off
Example: opus46?reasoning=auto		 — web_search (web_search_20260209)
web_fetch (web_fetch_20260209)
beta: code-execution-web-tools-2026-02-09
opus47 anthropic Text, Vision, Document json (schema) effort: auto, low, medium, high, xhigh, max, off
Example: opus47?reasoning=auto		 — web_search (web_search_20260209)
web_fetch (web_fetch_20260209)
beta: code-execution-web-tools-2026-02-09
opus4 anthropic Text, Vision, Document json (schema) effort: auto, low, medium, high, xhigh, max, off
Example: opus4?reasoning=auto		 — web_search (web_search_20260209)
web_fetch (web_fetch_20260209)
beta: code-execution-web-tools-2026-02-09
opus5 anthropic Text, Vision, Document json (schema) effort: auto, low, medium, high, xhigh, max, off
Example: opus5?reasoning=auto		 — web_search (web_search_20260209)
beta: code-execution-web-tools-2026-02-09
opus anthropic Text, Vision, Document json (schema) effort: auto, low, medium, high, xhigh, max
Example: opus?reasoning=medium		 — web_search (web_search_20260209)
beta: code-execution-web-tools-2026-02-09
sonnet4 anthropic Text, Vision, Document json (schema) effort: auto, low, medium, high, max, off
Example: sonnet4?reasoning=auto		 — web_search (web_search_20260209)
web_fetch (web_fetch_20260209)
beta: code-execution-web-tools-2026-02-09
astra codexresponses Text, Vision, Document json (schema) effort: low, medium, high, xhigh, max
Example: astra?reasoning=medium		 low, medium, high
Example: astra?verbosity=low		 —
codexspark codexresponses Text json (schema) — — —
luna56 codexresponses Text, Vision, Document json (schema) effort: none, low, medium, high, xhigh, max, off
Example: luna56?reasoning=high		 low, medium, high
Example: luna56?verbosity=low		 —
luna codexresponses Text, Vision, Document json (schema) effort: none, low, medium, high, xhigh, max, off
Example: luna?reasoning=medium		 low, medium, high
Example: luna?verbosity=low		 —
sol56 codexresponses Text, Vision, Document json (schema) effort: none, low, medium, high, xhigh, max, off
Example: sol56?reasoning=high		 low, medium, high
Example: sol56?verbosity=low		 —
sol codexresponses Text, Vision, Document json (schema) effort: none, low, medium, high, xhigh, max, off
Example: sol?reasoning=medium		 low, medium, high
Example: sol?verbosity=low		 —
terra codexresponses Text, Vision, Document json (schema) effort: none, low, medium, high, xhigh, max, off
Example: terra?reasoning=high		 low, medium, high
Example: terra?verbosity=low		 —
deepseek-v4-flash-vision-exp deepseek Text, Vision json (schema) effort: none, low, high, max, off
Example: deepseek-v4-flash-vision-exp?reasoning=max		 — —
deepseek-v4-flash deepseek Text, Vision json (schema) effort: none, low, high, max, off
Example: deepseek-v4-flash?reasoning=max		 — —
deepseek deepseek Text, Vision json (schema) effort: none, low, high, max, off
Example: deepseek?reasoning=max		 — —
deepseekpro deepseek Text json (schema) effort: none, low, high, max, off
Example: deepseekpro?reasoning=max		 — —
passthrough fast-agent Text json (schema) — — —
playback fast-agent Text json (schema) — — —
silent fast-agent Text json (schema) — — —
slow fast-agent Text json (schema) — — —
gemini25 google Text, Vision, Document, Audio, Video json (schema) effort: auto, minimal, low, medium, high, off
Example: gemini25?reasoning=auto		 — —
gemini25pro google Text, Vision, Document, Audio, Video json (schema) effort: auto, minimal, low, medium, high, off
Example: gemini25pro?reasoning=auto		 — —
gemini2 google Text, Vision, Document, Audio, Video json (schema) — — —
gemini3.1 google Text, Vision, Document, Audio, Video json (schema) effort: minimal, low, medium, high
Example: gemini3.1?reasoning=medium		 — —
gemini3.1flashlite google Text, Vision, Document, Audio, Video json (schema) effort: minimal, low, medium, high
Example: gemini3.1flashlite?reasoning=medium		 — —
gemini35 google Text, Vision, Document, Audio, Video json (schema) effort: minimal, low, medium, high
Example: gemini35?reasoning=medium		 — —
gemini37 google Text, Vision, Document, Audio, Video json (schema) effort: low, medium, high
Example: gemini37?reasoning=medium		 — —
gemini3 google Text, Vision, Document, Audio, Video json (schema) effort: minimal, low, medium, high
Example: gemini3?reasoning=medium		 — —
gemini3flash google Text, Vision, Document, Audio, Video json (schema) effort: minimal, low, medium, high
Example: gemini3flash?reasoning=medium		 — —
gemini google Text, Vision, Document, Audio, Video json (schema) effort: low, medium, high
Example: gemini?reasoning=medium		 — —
moonshotai/kimi-k2-instruct-0905 groq Text json (schema) — — —
moonshotai/kimi-k2-thinking groq Text json (schema) — — —
moonshotai/kimi-k2 groq Text json (schema) — — —
qwen3-32b groq Text json (object) — — —
qwen3.6-27b groq Text, Vision json (object) toggle: on, off
Example: qwen3.6-27b?reasoning=off		 — —
DeepSeek V4 Flash 0731 (baseten) hf Text json (schema) effort: none, low, high, max, off
Example: DeepSeek V4 Flash 0731 (baseten)?reasoning=max		 — —
Kimi K3 (together) hf Text, Vision json (schema) effort: low, high, max
Example: Kimi K3 (together)?reasoning=max		 — —
deepseek-ai/deepseek-v3.1 hf Text json (schema) — — —
deepseek-hf hf Text json (schema) — — —
deepseek32 hf Text json (schema) — — —
deepseek41-hf hf Text, Vision json (schema) effort: none, low, high, max, off
Example: deepseek41-hf?reasoning=max		 — —
gemma4 hf Text, Vision json (schema) effort: none, low, medium, high, off
Example: gemma4?reasoning=none		 — —
glimmer hf Text, Vision — effort: low, medium, high, xhigh
Example: glimmer?reasoning=high		 — —
glm47 hf Text json (schema) toggle: on, off
Example: glm47?reasoning=off		 — —
glm51 hf Text json (schema) toggle: on, off
Example: glm51?reasoning=off		 — —
glm5 hf Text json (schema) toggle: on, off
Example: glm5?reasoning=off		 — —
glm hf Text json (object) effort: none, minimal, low, medium, high, xhigh, max, off
Example: glm?reasoning=max		 — —
gpt-oss-20b hf Text json (schema) — — —
gpt-oss hf Text json (schema) — — —
hf.minimaxai/minimax-m2 hf Text json (schema) — — —
hf.qwen/qwen3-next-80b-a3b-instruct hf Text json (schema) — — —
hf.zai-org/glm-4.6 hf Text json (schema) — — —
kimi25 hf Text, Vision json (schema) toggle: on, off
Example: kimi25?reasoning=off		 — —
kimi26 hf Text, Vision json (schema) toggle: on, off
Example: kimi26?reasoning=off		 — —
kimi hf Text, Vision json (schema) — — —
minimax21 hf Text json (schema) — — —
minimax25 hf Text json (schema) toggle: on, off
Example: minimax25?reasoning=off		 — —
minimax27 hf Text json (schema) — — —
minimax hf Text, Vision json (schema) — — —
qwen/qwen3.8-27b hf Text, Vision, Video json (object) effort: low, medium, xhigh, off
Example: qwen/qwen3.8-27b?reasoning=medium		 — —
qwen35 hf Text, Vision json (object) toggle: on, off
Example: qwen35?reasoning=off		 — —
qwen36 hf Text — toggle: on, off
Example: qwen36?reasoning=off		 — —
zai-org/glm-5.3-flash hf Text, Vision, Document, Video json (object) effort: low, high, max
Example: zai-org/glm-5.3-flash?reasoning=max		 — —
zai-org/glm-5.3 hf Text json (object) effort: low, high, max
Example: zai-org/glm-5.3?reasoning=max		 — —
Muse Spark 1.1 metaai Text, Vision, Document, Video json (schema) effort: minimal, low, medium, high, xhigh
Example: Muse Spark 1.1?reasoning=medium		 — —
Muse Spark 1.2 (Contributor) metaai Text, Vision, Document, Video json (schema) effort: minimal, low, medium, high, xhigh
Example: Muse Spark 1.2 (Contributor)?reasoning=medium		 — —
Muse Spark 1.2 metaai Text, Vision, Document, Video json (schema) effort: minimal, low, medium, high, xhigh
Example: Muse Spark 1.2?reasoning=medium		 — —
muse metaai Text, Vision, Document, Video json (schema) effort: minimal, low, medium, high, xhigh
Example: muse?reasoning=medium		 — —
musecontrib metaai Text, Vision, Document, Video json (schema) effort: minimal, low, medium, high, xhigh
Example: musecontrib?reasoning=medium		 — —
kimik3 moonshot Text, Vision, Video json (schema) effort: low, high, max
Example: kimik3?reasoning=max		 — —
gpt-4.1-2025-04-14 openai Text, Vision, Document json (schema) — — —
gpt-4.1-mini-2025-04-14 openai Text, Vision, Document json (schema) — — —
gpt-4.1-mini openai Text, Vision, Document json (schema) — — —
gpt-4.1-nano-2025-04-14 openai Text, Vision, Document json (schema) — — —
gpt-4.1-nano openai Text, Vision, Document json (schema) — — —
gpt-4.1 openai Text, Vision, Document json (schema) — — —
gpt-4o-2024-11-20 openai Text, Vision, Document json (schema) — — —
gpt-4o-mini-2024-07-18 openai Text, Vision, Document json (schema) — — —
gpt-4o-mini openai Text, Vision, Document json (schema) — — —
gpt-4o openai Text, Vision, Document json (schema) — — —
chatgpt responses Text, Vision, Document json (schema) — — —
codex responses Text, Vision, Document json (schema) effort: low, medium, high, xhigh
Example: codex?reasoning=medium		 low, medium, high
Example: codex?verbosity=low		 —
gpt-5-mini responses Text, Vision, Document json (schema) effort: minimal, low, medium, high
Example: gpt-5-mini?reasoning=medium		 low, medium, high
Example: gpt-5-mini?verbosity=low		 —
gpt-5-nano-2025-08-07 responses Text, Vision, Document json (schema) effort: minimal, low, medium, high
Example: gpt-5-nano-2025-08-07?reasoning=medium		 low, medium, high
Example: gpt-5-nano-2025-08-07?verbosity=low		 —
gpt-5-nano responses Text, Vision, Document json (schema) effort: minimal, low, medium, high
Example: gpt-5-nano?reasoning=medium		 low, medium, high
Example: gpt-5-nano?verbosity=low		 —
gpt-5.3-chat-latest responses Text, Vision, Document json (schema) — — —
gpt-5 responses Text, Vision, Document json (schema) effort: minimal, low, medium, high
Example: gpt-5?reasoning=medium		 low, medium, high
Example: gpt-5?verbosity=low		 —
gpt51 responses Text, Vision, Document json (schema) effort: none, low, medium, high, xhigh, off
Example: gpt51?reasoning=medium		 low, medium, high
Example: gpt51?verbosity=low		 —
gpt52 responses Text, Vision, Document json (schema) effort: none, low, medium, high, xhigh, off
Example: gpt52?reasoning=medium		 low, medium, high
Example: gpt52?verbosity=low		 —
gpt54-mini responses Text, Vision json (schema) effort: none, low, medium, high, xhigh, off
Example: gpt54-mini?reasoning=medium		 low, medium, high
Example: gpt54-mini?verbosity=low		 —
gpt54-nano responses Text, Vision json (schema) effort: none, low, medium, high, xhigh, off
Example: gpt54-nano?reasoning=medium		 low, medium, high
Example: gpt54-nano?verbosity=low		 —
gpt54 responses Text, Vision, Document json (schema) effort: none, low, medium, high, xhigh, off
Example: gpt54?reasoning=medium		 low, medium, high
Example: gpt54?verbosity=low		 —
gpt55 responses Text, Vision, Document json (schema) effort: none, low, medium, high, xhigh, off
Example: gpt55?reasoning=medium		 low, medium, high
Example: gpt55?verbosity=low		 —
gpt56 responses Text, Vision, Document json (schema) effort: none, low, medium, high, xhigh, max, off
Example: gpt56?reasoning=high		 low, medium, high
Example: gpt56?verbosity=low		 —
o1-2024-12-17 responses Text, Vision json (schema) effort: minimal, low, medium, high, xhigh
Example: o1-2024-12-17?reasoning=medium		 — —
o1-mini responses Text, Vision json (schema) effort: minimal, low, medium, high, xhigh
Example: o1-mini?reasoning=medium		 — —
o1-preview responses Text, Vision json (schema) effort: minimal, low, medium, high, xhigh
Example: o1-preview?reasoning=medium		 — —
o1 responses Text, Vision json (schema) effort: minimal, low, medium, high, xhigh
Example: o1?reasoning=medium		 — —
o3-2025-04-16 responses Text, Vision, Document json (schema) effort: low, medium, high
Example: o3-2025-04-16?reasoning=medium		 — —
o3-mini-2025-01-31 responses Text json (schema) effort: low, medium, high
Example: o3-mini-2025-01-31?reasoning=medium		 — —
o3-mini responses Text json (schema) effort: low, medium, high
Example: o3-mini?reasoning=medium		 — —
o3 responses Text, Vision, Document json (schema) effort: low, medium, high
Example: o3?reasoning=medium		 — —
o4-mini-2025-04-16 responses Text, Vision, Document json (schema) effort: low, medium, high
Example: o4-mini-2025-04-16?reasoning=medium		 — —
o4-mini responses Text, Vision, Document json (schema) effort: low, medium, high
Example: o4-mini?reasoning=medium		 — —
responses.o3-pro responses Text json (schema) — — —
grok43 xai Text, Vision json (schema) effort: low, medium, high
Example: grok43?reasoning=high		 — —
grok45 xai Text, Vision json (schema) effort: low, medium, high
Example: grok45?reasoning=high		 — —
grok46 xai Text, Vision json (schema) effort: low, medium, high, xhigh
Example: grok46?reasoning=high		 — —
grok xai Text, Vision json (schema) effort: low, medium, high, xhigh
Example: grok?reasoning=high		 — —
grokfast xai Text, Vision json (schema) effort: low, medium, high, xhigh
Example: grokfast?reasoning=high		 — —
zaiglm53 zai Text json (object) effort: low, high, max
Example: zaiglm53?reasoning=max		 — —
zaiglm53flash zai Text, Vision, Document, Video json (object) effort: low, high, max
Example: zaiglm53flash?reasoning=max		 — —
zaiglm zai Text json (object) effort: none, minimal, low, medium, high, xhigh, max, off
Example: zaiglm?reasoning=max		 — —