OpenCode Go lists Ox Alpha Free as a limited-time model

OpenCode Go now lists Ox Alpha Free as a limited-time model. The live route changed from x-preview-f-free to ox-alpha-free during the same monitoring window, while the model remains a stealth release with no public developer or model card.
AI Neural Narration
48kHz StudioFish Audio Neural Engine · Natural editorial narration
Key Takeaways
- check_circleOpenCode officially lists Ox Alpha Free on Go as a free model for a limited time.
- check_circleAIMI saw the Go route under x-preview-f-free at 10:19:21 SAST, then saw that identifier removed and ox-alpha-free added 17 minutes and 26 seconds later.
- check_circleOpenCode calls Ox Alpha Free a stealth model. Its official announcement claims a 1M-token context, multimodal input and zero data retention, but does not identify the developer, weights or benchmark sheet.
What OpenCode has confirmed
OpenCode's Go page now names Ox Alpha Free in the model list and labels it as available for a limited time. Its Go documentation also describes the route as a free model and links it to the OpenCode Go API.
That confirms a provider availability change. It does not establish a separate model launch by xAI, OpenAI or another named lab. OpenCode's documentation calls the model stealth, so the underlying developer is not identified.
A route rename, not a second model release
The AIMI endpoint monitor first saw x-preview-f-free on OpenCode Go at 10:19:21 SAST. At 10:36:47 SAST, the same Go endpoint removed x-preview-f-free and added ox-alpha-free. The gap was 17 minutes and 26 seconds.
OpenCode's current page still documents x-preview-f-free as the API model ID, while the live endpoint now returns ox-alpha-free. For new tests, use the live endpoint as the availability check and expect the documentation to catch up with the rename.
What remains unknown
There is no public model card, weights repository or benchmark sheet attached to the Ox Alpha Free listing. OpenCode's announcement claims a 1M-token context and multimodal input, but it does not identify the developer. The name alone is not evidence that it comes from xAI, despite the earlier x-preview identifier.
OpenCode also says the provider follows a zero-retention policy and does not use requests for model training. That is a provider statement, not an independent audit, and the limited-time route can change without notice.
How to test it safely
Treat ox-alpha-free as an experimental provider route. Check https://opencode.ai/zen/go/v1/models before each test, keep a fallback model ready, and avoid sending confidential material until you have reviewed the current provider terms.
The route is useful for comparing agent behaviour during its free window, but there is not enough public information to make a capability or production-reliability claim. Record the exact route ID and date with any evaluation results.
Frequently Asked Questions
Is Ox Alpha Free an xAI model?
There is no public evidence confirming that. OpenCode calls it a stealth model, and the earlier x-preview-f-free route does not identify a developer.
Is Ox Alpha Free free to use?
OpenCode lists it as free on Go for a limited time. The Go subscription and its limits are separate from the model route, so check the current OpenCode terms before relying on it.
Which model ID should I use?
The live Go endpoint currently returns ox-alpha-free. OpenCode's documentation still shows x-preview-f-free, so check the endpoint immediately before configuring a client.
Related Frontier Models & Releases
Explore verified specifications, benchmark results, and route pricing across alternative models in this class.
DeepSeek releases V4.1 Flash with native multimodal vision, 552B MoE and lower API rates
DeepSeek has launched DeepSeek-V4.1-Flash, a 552B-parameter mixture-of-experts model featuring a novel Causal Encoder–Decoder architecture with just 8B active input and 16B active output parameters. The release brings native vision, compresses KV cache storage by up to 8x, and slashes off-peak API pricing to $0.15 per million input tokens.
OpenAI Releases GPT-6 Astra: Next-Generation Flagship with 1.05M Context and Deep Multimodal Reasoning
OpenAI has officially launched GPT-6 Astra, its frontier flagship model featuring a 1,050,000-token context window, 128,000 max output tokens, and native tool-use for autonomous agent workflows.
Alibaba Qwen Releases Qwen3.8 Max: 2.4-Trillion Parameter Flagship with 1M Multimodal Context
Alibaba's Qwen team has launched Qwen3.8 Max, a 2.4T parameter Mixture-of-Experts model offering 1M token context, native video perception, and deep agent tool orchestration.