Long-context guide

Best Long-Context Models

For giant documents, repos, transcripts, and agent memory buffers, context window only helps when it is paired with retrieval discipline and pricing you can actually afford.

ModelContextBest forInput $/M
GPT-6 Astra
OpenAI
1.05MHard end-to-end agent work$10
GPT-5.6 Sol
OpenAI
1.05MComplex production workflows$4
GPT-5.6 Terra
OpenAI
1.05MProduction agents$2
GPT-5.6 Luna
OpenAI
1.05MHigh-volume workflows$0.2
Gemini 3.8 Flash
Google
1.05MLong-horizon software engineering$0.75
Gemini 3.7 Flash
Google
1.05MAgentic coding$0.75
GPT-5.5
OpenAI
1MComplex reasoning$5
Claude Fable 5.1
Anthropic
1MDemanding reasoning$10
Claude Fable 5
Anthropic
1MLong-running agents$10
Claude Opus 5
Anthropic
1MComplex reasoning$5
Claude Opus 4.8
Anthropic
1MComplex reasoning$5
GPT-5.4
OpenAI
1MCoding$2.5
Gemini 3.5 Flash
Google
1MFast multimodal agents$1.5
Gemini 3.6 Flash
Google
1MAgentic coding$1.5
Gemini 3.5 Flash-Lite
Google
1MHigh-volume automation$0.3
Claude Sonnet 5
Anthropic
1MBalanced performance$2
GLM-5.3-Flash
Z.ai
1MCost-sensitive multimodal coding$0.15
GLM-5.3
Z.ai
1MComplex software engineering$1.4
DeepSeek V4 Pro
DeepSeek
1MBudget coding$0.44
DeepSeek V4 Flash
DeepSeek
1MLow-cost agent experiments$0.14
Kimi K3
Moonshot AI
1MLong-horizon coding$3
Grok 4.5
xAI
500KAgentic coding$2
Grok 4.6
xAI
500KAgentic coding$2
GPT-5.2-Codex
OpenAI
400KCoding-focused tasks$1.75
GPT-5.2
OpenAI
400KGeneral-purpose$1.75
Mistral Medium 3.5
Mistral
256KEuropean compliance$1.5
Claude Haiku 4.5
Anthropic
200KFast responses$1
GLM-5
Zhipu AI
200KBilingual (CN/EN)$1
GPT-OSS-120B
OpenAI
128KSelf-hosted$0
North Small Translate
Cohere
16KMachine translation$

Fast picks