GPT-5.3 Codex (xhigh)
Intelligence
54
#29/556
Coding
53.1
#165/163
Math
—
Speed
66 tok/s
TTFT: 72.24s
Pricing
$1.75 / $14.00
per 1M tokens (in/out)
GPT-5.3-Codex is OpenAI’s most capable coding model, combining frontier agentic coding capabilities, improvements in aesthetics, and context compaction. It sets new state-of-the-art results on Terminal-Bench 2.0 (77.3%), OSWorld-Verified (64.7%), and SWE-Lancer IC Diamond (81.4%). First model classified as High capability for cybersecurity under OpenAI’s Preparedness Framework. Available in the Codex app and API.
GPT-5.3 Codex (xhigh) is OpenAI’s advanced language model designed for coding tasks. It processes at 66.352 tokens per second and is priced at $1.75 per million input tokens, targeting professional developers and technical users.
When to Use GPT-5.3 Codex (xhigh)
✓ Best For
- Automating code generation
- Debugging and code review
- Providing coding assistance in IDEs
✗ Not Ideal For
- Users seeking general conversational AI
- Tasks requiring extensive mathematical computations
How GPT-5.3 Codex (xhigh) Compares
Intelligence Index · Higher is better
GoogleMiniMaxOpenAIDeepSeekKimi
Benchmark Profile
Output Speed · tok/s
AlibabaOpenAIMetaXiaomi
Intelligence · Coding · Math
Intelligence
Coding
Math
All Benchmark Scores (16)
| Benchmark | Score | |
|---|---|---|
| Intelligence Index | 54 | |
| Coding Index | 53.1 | |
| GPQA | 915% | |
| HLE | 399% | |
| SciCode | 53.2% | |
| IFBench | 75.4% | |
| LCR | 74% | |
| TerminalBench Hard | 53% | |
| Tau2 | 86% | |
| Cybersecurity CTFs | 77.6% | |
| OSWorld-Verified | 64.7% | |
| SWE-Bench Pro | 56.8% | |
| SWE-Lancer (IC-Diamond subset) | 81.4% | |
| Terminal-Bench 2.0 | 77.3% | |
| Arena Chat | 21.5 | |
| Arena Coding | 13.9 |
Data: Artificial Analysis · Updated: April 10, 2026
Frequently Asked Questions (15)
When was GPT-5.3 Codex (xhigh) released?
GPT-5.3 Codex (xhigh) was released on February 5, 2026.
Who created GPT-5.3 Codex (xhigh)?
GPT-5.3 Codex (xhigh) was created by OpenAI.
How intelligent is GPT-5.3 Codex (xhigh)?
GPT-5.3 Codex (xhigh) scores 54 on the Artificial Analysis Intelligence Index, placing it well above average among other reasoning models in a similar price tier (median: 31).
How fast is GPT-5.3 Codex (xhigh)?
GPT-5.3 Codex (xhigh) generates output at 68.8 tokens per second (based on OpenAI's API), which is above average compared to other reasoning models in a similar price tier (median: 67.8 t/s).
What is the latency of GPT-5.3 Codex (xhigh)?
GPT-5.3 Codex (xhigh) has a time to first token (TTFT) of 95.74s (based on OpenAI's API), which is at the higher end compared to other reasoning models in a similar price tier (median: 2.64s).
How much does GPT-5.3 Codex (xhigh) cost?
GPT-5.3 Codex (xhigh) costs $1.75 per 1M input tokens (somewhat higher than average, median: $1.35) and $14.00 per 1M output tokens (somewhat higher than average, median: $8.40), based on OpenAI's API.
What is GPT-5.3 Codex (xhigh) API pricing?
GPT-5.3 Codex (xhigh) costs $1.75 per 1M input tokens and $14.00 per 1M output tokens (based on OpenAI's API). For a blended rate (3:1 input to output ratio), this is $4.81 per 1M tokens. Pricing may vary by provider.
How verbose is GPT-5.3 Codex (xhigh)?
When evaluated on the Intelligence Index, GPT-5.3 Codex (xhigh) generated 77M output tokens, which is at the higher end compared to other reasoning models in a similar price tier (median: 30M).
Is GPT-5.3 Codex (xhigh) a reasoning model?
Yes, GPT-5.3 Codex (xhigh) is a reasoning model. It uses extended thinking or chain-of-thought reasoning to work through complex problems before providing an answer.
What input modalities does GPT-5.3 Codex (xhigh) support?
GPT-5.3 Codex (xhigh) supports text and image input.
What output modalities does GPT-5.3 Codex (xhigh) support?
GPT-5.3 Codex (xhigh) supports text output.
Can GPT-5.3 Codex (xhigh) process images?
Yes, GPT-5.3 Codex (xhigh) supports image input and can analyze, describe, and answer questions about images.
Is GPT-5.3 Codex (xhigh) multimodal?
Yes, GPT-5.3 Codex (xhigh) is multimodal. It can process text and image input and generate text output.
What is the context window of GPT-5.3 Codex (xhigh)?
GPT-5.3 Codex (xhigh) has a context window of 400k tokens. This determines how much text and conversation history the model can process in a single request.
Is GPT-5.3 Codex (xhigh) open source?
No, GPT-5.3 Codex (xhigh) is proprietary. The model weights are not publicly available.
