Add corrections, implementation notes, pricing changes, or usage caveats for other readers.
Dec 11, 2025
Knowledge cutoff
2025-08-31
Input modalities
Output modalities
Capabilities
128,000 tokens
Context window
400,000 tokens
Recent tweets and retweets from AIHubMix
Kimi K3 vs Opus 4.8 vs GPT-5.6 Sol
One Fruit Ninja prompt. One generation per model.
Claude Opus 4.8:3m46s · $0.554
GPT-5.6 Sol :12m01s · $0.392
Kimi K3 : 42m21s · $1.131
K3 used ~77K tokens, including ~54K reasoning tokens.
This is only one sample, primarily visual…
Use it now⬇️
aihubmix.com/model/kimi-k3
Link
kimi-k3 - API Pricing & Performance | AIHubMix
Kimi K3 is Kimi’s flagship model for long-horizon coding and… $3 per million input tokens, $15 per million output tokens.
aihubmix.com
🚀 Kimi K3 is now live on AiHubMix.
@Kimi_Moonshot AI's most capable model yet: 2.8T params, built on Kimi Delta Attention and Attention Residuals.
-1M token context
-Native multimodal (text + image → text)
-Built for agentic coding, tool use & debugging across huge…
gpt-4o-transcribe-diarize: An automatic speech recognition (ASR) model with built-in speaker diarization, meaning it associates audio segments with different speakers in a conversation. This model is only available in the Transcription API.
Use it…
gpt-audio-1.5: OpenAI’s first generally available audio model. It accepts audio inputs and outputs, and can be used in the Chat Completions REST API.
Use it now:
aihubmix.com/model/gpt-audio…
Link
AIHubMix - One Interface, Router All LLMs
Access every major LLM through a…
Discuss this model
Add corrections, implementation notes, pricing changes, or usage caveats for other readers.