GPT-5.3 Codex
LLM Provider / Model · Agentic coding across the Codex app, CLI, IDE and web
At a glance
| Input price | $1.75 |
| Output price | $14 |
| Cache price | $0.175 |
| Price tier | Mid |
| Context window | 400K |
| Max output | 128K |
| Context tier | 128K-500K |
| Speed tier | Slow/Reasoning |
| Latency | high |
| Knowledge cutoff | Aug 2025 |
| Modality | Multimodal (vision) |
| Model ID | gpt-5.3-codex |
| Provider | OpenAI |
| Benchmark | Purpose-built for agentic coding rather than general chat |
| Released | Feb 2026 |
| Hosting | Closed/API |
| Capabilities | Agentic coding, Steerable long-running tasks, Responses API, Vision, Tool use, Structured output |
What GPT-5.3 Codex does
Agentic coding, Steerable long-running tasks, Responses API, Vision, Tool use, Structured output
Best for
Agentic coding across the Codex app, CLI, IDE and web
Works well with
Integration
Agent / Orchestration
Conflicts & caveats
- Kiro's spec mode is tuned for Claude models — GPT-5.3 Codex may work but spec-driven flows are best with Claude.
- Gemini CLI is locked to Google Gemini models — your LLM is GPT-5.3 Codex.
- Claude Code only works with Claude models — your LLM is GPT-5.3 Codex.
- Devin uses Cognition's proprietary model internally — your selected GPT-5.3 Codex will be ignored.
- OpenAI Codex (Cloud) runs GPT-5.1 Codex internally — your selected GPT-5.3 Codex will be ignored.
- Jules runs Gemini models internally — your selected GPT-5.3 Codex will be ignored.
- The Claude Code SDK is tuned for Claude models — GPT-5.3 Codex may work via the model adapter but loses Claude-specific features (thinking, prompt caching, sub-agents).
- Lovable uses a fixed internal model — your selected GPT-5.3 Codex will be ignored.
- Replit Agent 3 uses its own model selection internally — your selected GPT-5.3 Codex will be ignored.
Build a full stack around GPT-5.3 Codex — Flowpicker shows compatibility warnings before you commit.
Open the stack planner →