Kimi K3 (batch)
Kimi K3 · Batch tier (Batch)
Asynchronous batch tier: discounted offline batch requests with high throughput and relaxed latency — ideal for large-scale non-real-time workloads.
Best for: Large-scale offline inference, cost-sensitive batch jobs, backfills
This page covers the Kimi K3 Batch tier, sharing the underlying model capability with its real-time standard tier; the main difference is in billing mode and invocation method. See the Kimi K3 standard real-time tier.
Provider：Moonshot AI
Kimi K3 is a 2.8T parameter open-weight multimodal reasoning model from Moonshot AI. It is suited for complex coding, knowledge work, and long-horizon agentic workflows, and is particularly strong at...
Key specs
- Context：1M
- Max output：943.7K
- Tokenizer：Other
- Released：2026-07-16
Token pricing
- Input price：$3.00 / 1M tokens
- Output price：$15.00 / 1M tokens
- Cache read：$0.300 / 1M tokens
- Blended price：$6.00 / 1M tokens
Modalities
Text, Image, Video
Use Cases
- Deep Reasoning & Analysis：Process long legal docs, academic papers, financial reports; multi-hop reasoning and evidence-chain tracing across huge contexts.（≈ $0.120/call）
- Vision Understanding & Generation：Recognize data trends in charts, OCR scanned docs, analyze photos and produce structured reports.（≈ $0.090/call）
When to pick this model
Signals derived from public pricing and spec fields — not a benchmark. Validate with official docs and your own eval.
- Long context (>=1M tokens): well-suited for long-doc QA, code-base analysis, multi-turn agents.
- Long max output (>=30K tokens): strong fit for deep report / essay / document-drafting workflows.
- Multimodal (text+image): fit for image captioning, OCR, chart QA, vision-grounded reasoning.
- Video input support: fit for video understanding, frame QA, content moderation.
FAQ
- What is the token pricing of Kimi K3 (batch)? Input price $3.00 / 1M tokens, output price $15.00 / 1M tokens, blended about $6.00 / 1M tokens. Refer to Moonshot AI’s official page for the exact rate.
- What context window does Kimi K3 (batch) support? Context window is 1M, max output about 943.7K, suitable for long documents, multi-turn chat and complex reasoning.
- What scenarios is Kimi K3 (batch) best for? Based on its capability and pricing, Kimi K3 (batch) fits best: Deep Reasoning & Analysis, Vision Understanding & Generation. See the Use Cases section for details and cost estimates.
- What input/output modalities does Kimi K3 (batch) support? Supports Text, Image, Video modality.
- Is Kimi K3 (batch) a free model? Kimi K3 (batch) is billed per token, not a free model.
- Which provider offers Kimi K3 (batch)? Kimi K3 (batch) is offered by Moonshot AI.
More models from Moonshot AI
Full info：Kimi K3 (batch)