Sulat.com
AI models
$10 off the fastest DeepSeek V4.1 Flash, Kimi K3 and GLM 5.3
$10 off the fastest DeepSeek V4.1 Flash, Kimi K3 and GLM 5.3 from Synthetic
Poe logo

Model details

GPT-5.3-Codex-Spark

GPT-5.3-Codex-Spark is a specialized, smaller-scale iteration of the GPT-5.3-Codex architecture, engineered specifically to prioritize responsiveness in software development environments. While many AI assistants function as external chat interfaces, this model is built to integrate directly into the developer's workflow, allowing for immediate, targeted edits and rapid logic adjustments. By focusing on low-latency performance, it enables a more fluid, iterative experience where developers can refine interfaces and debug code with near-instant feedback, bridging the gap between autonomous, long-running tasks and the need for immediate, hands-on control.

Developed as the first milestone in a strategic partnership with Cerebras, the model leverages specialized wafer-scale hardware to achieve speeds exceeding 1,000 tokens per second. This optimization ensures that the model remains highly capable for real-world coding tasks while maintaining the high-speed throughput necessary for live, in-editor interaction. As a research preview, it serves as a testing ground for high-performance inference, allowing developers to experiment with rapid iteration cycles across large codebases and complex modules before the deployment of larger, more resource-intensive frontier models.

Poeopenai/gpt-5.3-codex-spark
Open provider pageProvider docs

Quick Info

Powered by
Provider
Poe
Model key
openai/gpt-5.3-codex-spark
Release date
Mar 4, 2026
Last updated
Mar 4, 2026
Input modalities
Output modalities
Capabilities

Cost

A provider subscription or plan supersedes token-based pricing for this model.

Limits

Output tokens
16,384 tokens
Context window
128,000 tokens
News search byUse it for free

Latest news about GPT-5.3-Codex-Spark

Poe

Coverage

BigGo Finance reports that OpenAI announced GPT-5.3-Codex-Spark would be retired roughly seven months after its February 12, 2026 launch, with Codex head Thibault Sottiaux announcing the decision on X. The model had debuted at over 1,000 tokens per second on Cerebras' WSE-3 wafer-scale chips and was framed as OpenAI's The article attributes the retirement to declining usage and the availability of clearly superior models, and cites Cerebras' August introduction of an Ultrafast mode running the flagship GPT-5.6 Sol at 750 tokens per second as eliminating Spark's core differentiator. It also notes a broader pattern of OpenAI retiring

Poe

Coverage

OpenAI's GPT-5.3-Codex-Spark, now on Cerebras hardware, delivers over 1,000 tokens/sec for real-time coding and pair-programming.

Poe

Coverage

OpenAI released the Codex-Spark coding model to help developers to make targeted edits, adjust logic, refine interfaces, and more.

Poe

Coverage

OpenAI on Thursday released GPT-5.3-Codex-Spark, its first AI model served on chips from Cerebras Systems, marking the ChatGPT maker’s first production deployment on non-Nvidia silicon.

Poe

Coverage

OpenAI released the ultra-fast coding AI model ' GPT-5.3-Codex-Spark ' on February 12, 2026. GPT-5.3-Codex-Spark runs on an AI chip developed by Cerebras , achieving both high processing performance and fast response times. Introducing GPT-5.3-Codex-Spark | OpenAI https://openai.com/index/introducing-gpt-5-3-codex-spar

Videos about GPT-5.3-Codex-Spark