The right model.
Always within reach.
Explore the curated AIx catalog across chat, reasoning, image generation, speech and embeddings. One OpenAI-compatible key calls every model below.
| Model | Best for | Type | Price | Unit |
|---|---|---|---|---|
| Claude Sonnet 4.6anthropic/claude-sonnet-4-6 | Balanced coding & agents — the default pick | chat | $7.5 | / 1M tokens |
| Claude Opus 4.8anthropic/claude-opus-4-8 | Hardest reasoning & long agentic sessions | chat | $12.5 | / 1M tokens |
| Claude Haiku 4.5anthropic/claude-haiku-4-5 | Fast agent turns and high-volume coding support | chat | $2.5 | / 1M tokens |
| Kimi K2.7 Codemoonshotai/Kimi-K2.7-Code | Agentic, tool-using software engineering | chat | $4 | / 1M tokens |
| DeepSeek V3.2deepseek-ai/DeepSeek-V3.2 | Strong open-model coding at very low cost | chat | $0.38 | / 1M tokens |
| DeepSeek V4 Flashdeepseek-ai/DeepSeek-V4-Flash | Fast general and coding workloads | chat | $0.2 | / 1M tokens |
| GLM 4.6zai-org/GLM-4.6 | Coding & agents with strong price/performance | chat | $1.74 | / 1M tokens |
| Kimi K2.6moonshotai/Kimi-K2.6 | Agentic workflows & heavy tool use | chat | $4.5 | / 1M tokens |
| DeepSeek R1 0528deepseek-ai/DeepSeek-R1-0528 | Open reasoning workloads | chat | $2.15 | / 1M tokens |
| Gemini 2.5 Progoogle/gemini-2.5-pro | Massive context & multimodal analysis | chat | $4.2969 | / 1M tokens |
| Gemini 2.5 Flashgoogle/gemini-2.5-flash | Fast, cheap, very long context | chat | $1.0625 | / 1M tokens |
| GPT OSS 120Bopenai/gpt-oss-120b | Open general-purpose reasoning | chat | $0.6 | / 1M tokens |
| Llama 3.3 70B Instruct Turbometa-llama/Llama-3.3-70B-Instruct-Turbo | Open general-purpose workhorse | chat | $1.04 | / 1M tokens |
| FLUX 1.1 Problack-forest-labs/FLUX-1.1-pro | High-fidelity image generation | image | $0.04 | / image |
| Kokoro 82Mhexgrad/Kokoro-82M | Natural, low-cost text-to-speech | tts | $0.62 | / 1M tokens |
| Whisper Large V3openai/whisper-large-v3 | Accurate multilingual transcription | audio | $0.0005 | / min |
| BGE Large EN v1.5BAAI/bge-large-en-v1.5 | General-purpose retrieval embeddings | embeddings | $0.01 | / 1M tokens |
Match the model
to the workload.
Start from the outcome you need, then balance quality, context, latency and unit economics. AIx keeps the integration consistent when your model choice changes.
Coding & agents
Models selected for code generation, repository reasoning, tools and multi-step execution.
Build a coding agent → 02 / REASONLong context
Understand context windows, output limits and cost before sending large documents.
Plan token usage → 03 / CREATEImages & media
Generate visual assets, speech and transcripts with modality-specific endpoints.
Explore media APIs → 04 / RETRIEVESearch & RAG
Create semantic search, retrieval pipelines and durable memory with embedding models.
Build retrieval →Everything you need
to choose and ship.
How do I call an AI model on AIx?+
Create one AIx API key, use the OpenAI-compatible base URL and send the canonical model id in your request.
Does every model use the same API key?+
Yes. The same scoped AIx key and prepaid balance work across supported chat, image, speech, transcription and embedding models.
How is AI model pricing calculated?+
Chat and embedding models are metered by tokens, image models per generated image and transcription models per minute. The live catalog is the pricing source of truth.