Models

List of available models with their features, capabilities, and API identifiers.

API Model ID is the exact string to use in your API requests, e.g. "model": "deepseek-v3.2"

FreeFree plan

Available to all users at no cost. Great for prototyping and lightweight workloads.

PaidPro or Premium plan

Available on any paid plan. Pro and Premium both unlock the full model catalog.

Expected latency: Most models respond in 1–3s. Large MoE models (GLM-5, Kimi K2 series) may take 3–6s due to model size. Reasoning models spend additional tokens on internal thinking — use max_tokens >= 500 to avoid truncation.

Text & Chat

Model Name	API Model ID	Tier	Primary Use Case	Rough Equivalent
DeepSeek V4 Pro	deepseek-v4-pro	Paid	Frontier-class text generation and reasoning	GPT-4o / Claude Opus tier
DeepSeek-R1-0528	deepseek-r1-0528	Paid	Frontier-class reasoning and analysis	GPT-4.1 / Claude Opus tier
GPT-OSS 120B	gpt-oss-120b	Paid	Complex reasoning and long-context tasks	GPT-4o tier
Kimi-K2-Thinking	kimi-k2-thinking	Paid	Deep reasoning and long-form analysis	GPT-4 reasoning tier
Kimi-K2.5	kimi-k2.5	Paid	Balanced reasoning with improved speed	Claude Sonnet tier
Llama 3.3 70B	llama-3.3-70b	Paid	High-quality chat and summarization	GPT-4-class tier
Qwen 3 32B	qwen-3-32b	Paid	Multilingual reasoning	Claude Sonnet tier
Kimi K2.6	kimi-k2.6	Paid	Multimodal reasoning with video input	Claude Opus 4.6 tier
Gemma 3 27B	gemma-3-27b	Paid	High-quality text generation	GPT-4o Mini tier
GLM 5	glm-5	Paid	Complex systems engineering and reasoning	Claude Opus 4.5 tier
Minimax M2.5	minimax-m2.5	Paid	Coding and agentic tool use	Claude Sonnet tier
Qwen 3 Coder 30B	qwen-3-coder-30b	Paid	Software development and technical reasoning	GPT-4-class coding
DeepSeek R1 70B	deepseek-r1-70b	Paid	Advanced reasoning, structured outputs	GPT-4-class reasoning tier
DeepSeek V3 0324	deepseek-v3-0324	Paid	Enhanced general-purpose chat	GPT-4o Mini tier
GPT-OSS 20B	gpt-oss-20b	Paid	Balanced chat and reasoning	GPT-4o Mini tier
Llama 3.1 8B	llama-3.1-8b	Paid	Fast general chat	GPT-4o Mini tier
Llama-4-Maverick-17B	llama-4-maverick-17b	Paid	Versatile MoE model for diverse tasks	GPT-4o tier
Ministral-14B	ministral-14b	Paid	Fast reasoning and structured outputs	GPT-4o Mini tier
DeepSeek R1 8B	deepseek-r1-8b	Free	Lightweight reasoning and chat	GPT-4o Mini tier
DeepSeek V3.2	deepseek-v3.2	Free	General-purpose chat and analysis	GPT-4o Mini / Claude Sonnet tier
DeepSeek V4 Flash	deepseek-v4-flash	Paid	1M context reasoning and coding	GPT-4o Mini / Claude Sonnet tier
Llama 3.2 3B	llama-3.2-3b	Free	Lightweight tasks and fast responses	GPT-3.5 tier
Mistral 7B	mistral-7b	Free	Fast lightweight generation	GPT-3.5 tier
Gemma 3 4B	gemma-3-4b	Free	Compact efficient inference	GPT-3.5 tier

Coding

Model Name	API Model ID	Tier	Primary Use Case	Rough Equivalent
DeepSeek Coder 33B	deepseek-coder-33b	Paid	Advanced code reasoning and generation	GPT-4-class coding
DeepSeek V4 Pro	deepseek-v4-pro	Paid	Frontier-class code reasoning and generation	GPT-4o / Claude Opus coding tier
DeepSeek V4 Flash	deepseek-v4-flash	Paid	1M context code reasoning and generation	GPT-4o Mini coding tier

Image Generation

Model Name	API Model ID	Tier	Primary Use Case	Rough Equivalent
Oxlo Image Pro	oxlo-image-pro	Paid	High-quality text-to-image	DALL·E 3 / Flux Pro tier
Oxlo Image Ultra	oxlo-image-ultra	Paid	Flagship photorealistic image generation	Midjourney v6 tier
Flux.1 Schnell	flux.1-schnell	Paid	Fast image generation	SD Turbo tier
SDXL Lightning	sdxl	Paid	Fast 4-step image generation	SDXL tier
Stable Diffusion 1.5	stable-diffusion-1.5	Free	Lightweight image generation	Early SD tier

Audio

Model Name	API Model ID	Tier	Primary Use Case	Rough Equivalent
Whisper Medium	whisper-medium	Free	Speech transcription	Whisper medium tier
Whisper Large	whisper-large	Free	High-accuracy transcription	Whisper large tier
Whisper Large v3	whisper-large-v3	Free	Production-grade transcription	Whisper large-v3 tier
Kokoro 82M	kokoro-82m	Free	Voice synthesis (TTS)	Standard TTS API tier

Embeddings

Model Name	API Model ID	Tier	Primary Use Case	Rough Equivalent
BGE-Large	bge-large	Free	Semantic search and RAG	text-embedding-large tier
E5-Large	e5-large	Free	Semantic clustering and retrieval	embedding-large tier

Computer Vision

Model Name	API Model ID	Tier	Primary Use Case	Rough Equivalent
YOLOv9	yolo-v9	Free	Real-time object detection	Real-time CV API tier
YOLOv11	yolo-v11	Free	Advanced object detection	Advanced CV detection tier

Rough Equivalent indicates general performance tier comparison and does not imply identical architecture or behavior.