GPT-4
Nov 2023
GPT-4Compare inon
Benchmarks of GPT-49
BenchmarkByScoreRank
Overall1
Knowledge4
Agents1
Writing2
Legacy · saturated or no longer reported1
No benchmark matches
Rankingin non-thinking mode?
Retired models are left out of the current rankings.
Alternatives?
ModelScoreCostPer taskWhat changes
CheaperClaude Opus 5.5max×8.6-79%$5.98No forced tool use+5BetterGPT-6 Lunalow×3.2-99%$0.005Function calling only with none · Chat Completions+3
OpenMiMo v2.6 Prothinking×6.9-99%$0.13Pass back reasoning_content+3CheaperClaude Sonnet 5.5max×8.4-89%$5.46No forced tool use+5CheaperClaude Fable 5.1max×8-47%$7.63Replace forced tool_choice+5ModelScoreCostPer taskWhat changes
Claude Opus 5.5max×8.6-79%$5.98No forced tool use+5Claude Sonnet 5.5max×8.4-89%$5.46No forced tool use+5Claude Fable 5.1max×8-47%$7.63Replace forced tool_choice+5GPT-6 Astramax×7.9-47%$3.26Monitoring can return HTTP 403+2
Gemini 4 Argonhigh×7.8-89%*$1.99Synchronous output stops at 65,536+2ModelScoreCostPer taskWhat changes
GPT-6 Lunalow×3.2-99%$0.005Function calling only with none · Chat Completions+3
Granite 4.2 3BthinkingOpen+35%-100%$0.006Gemma 4 E4BthinkingOpen+33%-100%$0.008GPT-5.6 Lunalow×3.1-99%$0.01GPT OSS 20BhighOpen+34%-100%$0.011ModelScoreCostPer taskWhat changes
MiMo v2.6 ProthinkingOpen×6.9-99%$0.13Pass back reasoning_content+3
DeepSeek V4.1 Flash 1017maxOpen×6.7-99%$0.26Pass back reasoning_content · Chat Completions+4
GLM 5.3maxOpen×6.7-94%$2.01Reasoning cannot be disabled+3
Step 5 PreviewthinkingOpen×6.5-96%$1.03web_search not available · Chat Completions+2
Kimi K3maxOpen×6.5-84%$2.00Sampling parameters are fixed+3ModelScoreCostPer taskWhat changes
GPT-6 Astramedium×7.4—$1.54Monitoring can return HTTP 403+2GPT-6.1 Solmedium×7.1—$0.21Remove unsupported parameters+5GPT-6 Solmedium×5.9—$0.25GPT-5.6 Solmedium×5.9—$0.50GPT-5.5medium×5—$0.90Prompts with >272K input tokensGPT-5.3 Codexxhigh×4.9—$1.75 / $14DeprecatedGPT-5.6 Terramedium×4.5—$0.18Prompts with >272K input tokensGPT-6 Lunamedium×4.5—$0.017Function calling only with none · Chat Completions+3All OpenAI models →History
17 versionsBenchmarks
all categoriesChat & Knowledge
AA Omniscience Index-33.8
GPQA Diamond34.9
Agents & Tool Use
Writing & Language
FAQ
How much does GPT-4 cost?›
$30 per 1M input tokens and $60 per 1M output tokens.
Is GPT-4 free to use?›
There is no free API; the cheapest paid price is $30 per 1M input tokens.
How can I use GPT-4?›
Through the API of Azure, OpenAI.
What is the context window of GPT-4?›
8K tokens, with up to 4K output tokens.
When was GPT-4 released? Is there a newer version?›
Released 2023-11-06. It has been superseded by GPT-5.6 Sol.
Is GPT-4 open source?›
No, GPT-4 is a closed model available through an API. The closest open-weights option here is MiMo v2.6 Pro.
Can GPT-4 read images?›
No, it accepts text only.
What is GPT-4?›
GPT-4 is a closed language model from OpenAI, released Nov 2023.
What are the best alternatives to GPT-4?›
Claude Opus 5.5 (cheaper), GPT-6 Luna (better), MiMo v2.6 Pro (open weights).