Skip to main content

    OpenAI

    GPT-5.3 Codex

    OpenAI's February 5, 2026 agentic coding model for building, debugging and maintaining software. At launch, OpenAI reported 56.8% on SWE-Bench Pro (Public) with xhigh reasoning effort. Accepts text and images and produces text, with a 400K-token context window, up to 128K output tokens and low, medium, high or xhigh reasoning effort.

    400K tokens
    Multimodal
    Max
    Provider
    OpenAI
    Context
    400K
    Catalog · As of Aug 11, 2026Stale · 31d old
    yno subscription tier
    Max
    Released
    Feb 5, 2026
    Speed
    Medium
    Reasoning
    Expert
    Modality
    Multimodal

    Benchmarks

    Quality

    AA Intelligence Index
    32.5points
    Artificial Analysis · Date unknownAge unknown
    GPQA Diamond
    91.5%
    Artificial Analysis · Date unknownAge unknown

    Speed and latency

    Output speed
    136.31tokens/s
    Configuration: Prompt length: 1,000 · Parallel queries: 1
    Artificial Analysis · Date unknownAge unknown
    Time to first token
    45.91s
    Configuration: Prompt length: 1,000 · Parallel queries: 1
    Artificial Analysis · Date unknownAge unknown
    Time to first answer token
    45.91s
    Configuration: Prompt length: 1,000 · Parallel queries: 1
    Artificial Analysis · Date unknownAge unknown

    API pricing

    API input cost
    $1.75/ 1M tokens
    Artificial Analysis · Date unknownAge unknown
    API output cost
    $14.00/ 1M tokens
    Artificial Analysis · Date unknownAge unknown

    Benchmark variant: GPT-5.3 Codex (xhigh)

    AA data retrieved Sep 11, 2026 · Artificial Analysis

    Indexes use points; evaluations use accuracy percentages. API measurements do not measure yno application performance. Methodology

    Capabilities

    • Code
    • Agentic Coding
    • Debugging
    • Refactoring
    • Tool Use
    • Vision

    Use cases

    • Autonomous coding agents
    • Large-scale refactoring
    • CI/CD automation
    • Code review

    Strengths

    • SWE-Bench Pro (Public) 56.8% at launch, xhigh
    • 400K context / 128K output
    • Text and image input
    • Agentic coding

    Best for

    Engineering teams building autonomous coding pipelines and CI/CD agents

    Use GPT-5.3 Codex in yno.ai

    No credit card required

    Related models

    Frequently asked

    What is GPT-5.3 Codex?
    OpenAI's February 5, 2026 agentic coding model for building, debugging and maintaining software. At launch, OpenAI reported 56.8% on SWE-Bench Pro (Public) with xhigh reasoning effort. Accepts text and images and produces text, with a 400K-token context window, up to 128K output tokens and low, medium, high or xhigh reasoning effort.
    How much does GPT-5.3 Codex cost in yno.ai?
    GPT-5.3 Codex is available on the Max tier of yno.ai.
    What can GPT-5.3 Codex do?
    GPT-5.3 Codex is best for Engineering teams building autonomous coding pipelines and CI/CD agents. Its main capabilities include Code, Agentic Coding, Debugging, Refactoring, Tool Use, Vision.
    How does GPT-5.3 Codex compare to other models?
    GPT-5.3 Codex excels at SWE-Bench Pro (Public) 56.8% at launch, xhigh and is recommended for Autonomous coding agents, Large-scale refactoring, CI/CD automation, Code review. See related models below.
    How do I use GPT-5.3 Codex in yno.ai?
    Sign up for yno.ai, select GPT-5.3 Codex from the model picker, and start chatting.