Cheapest LLM API for high-volume inference
Depot Pick: Fireworks AI. Runner-up: Together AI (then Groq). Job: High-volume cheap LLM inference. Updated 2026-11-04 (newest stored date on this page).
Millions of tokens/day for classification, extraction, or drafts where frontier quality is optional.
Ready call for agents
MCP https://getagentdepot.com/api/mcp → decide {"task":"cheap LLM API for high volume inference"}
HTTP GET /api/decide?jtbd=cheap-volume-llmNo key. The reply carries the pick, runner-up, the pick's sourced price with its date, and a recheck token.
Depot Pick: Fireworks AI
Editorial score 8.35 / 10 (curated estimate, as of 2026-10-28). Prices below are copied from the provider's own page; units are never converted.
pricePerf + reliability + dx. Open-model hosts win; Gemini Flash also excellent closed-model value.
| Item | Price (USD) | Source | Price as of | Last re-checked |
|---|---|---|---|---|
| Embeddings up to 150M params | $0.008 / million input tokens | official page | 2026-10-28 | not reconfirmed |
| Embeddings 150M-350M params | $0.016 / million input tokens | official page | 2026-10-28 | not reconfirmed |
| Qwen3 8B | $0.1 / million input tokens | official page | 2026-10-28 | 2026-11-04 |
| Fireworks Serverless: GLM 5.3 Flash Input ($0.15 / 1M input tokens) | $0.15 / million input tokens | official page | 2026-10-28 | not reconfirmed |
Runner-up 1: Together AI
Editorial score 8.20 / 10 (curated estimate, as of 2026-10-28). Prices below are copied from the provider's own page; units are never converted.
| Item | Price (USD) | Source | Price as of | Last re-checked |
|---|---|---|---|---|
| DeepSeek V4 Pro 0813 | $1.32 / million input tokens | official page | 2026-10-28 | 2026-11-04 |
| gpt-oss-120B | $0.15 / million input tokens | official page | 2026-10-28 | 2026-11-04 |
| GLM-5.3-Flash | $0.15 / million input tokens | official page | 2026-10-28 | 2026-11-04 |
| DeepSeek V4 Pro 0813 | $3.96 / million output tokens | official page | 2026-10-28 | 2026-11-04 |
Runner-up 2: Groq
Editorial score 7.95 / 10 (curated estimate, as of 2026-10-28). Prices below are copied from the provider's own page; units are never converted.
| Item | Price (USD) | Source | Price as of | Last re-checked |
|---|---|---|---|---|
| openai/gpt-oss-120b | $0.15 / million input tokens | official page | 2026-10-28 | 2026-11-04 |
| openai/gpt-oss-20b | $0.075 / million input tokens | official page | 2026-10-28 | 2026-11-04 |
| meta-llama/llama-prompt-guard-2-22m | $0.03 / million input tokens | official page | 2026-10-28 | 2026-11-04 |
| meta-llama/llama-prompt-guard-2-86m | $0.04 / million input tokens | official page | 2026-10-28 | 2026-11-04 |
Lowest sourced prices per million input tokens
Same unit only, price only (not quality-adjusted). The pick is editorial and is not chosen by price.
| Provider | Item | USD / million input tokens | Source | Price as of | Last re-checked |
|---|---|---|---|---|---|
| Cloudflare Workers AI | @cf/ibm-granite/granite-4.0-h-micro | $0.017 | official page | 2026-11-04 | 2026-11-04 |
| Chutes | Qwen3.8-27B (Qwen/Qwen3.8-27B-TEE) | $0.039 | official page | 2026-10-27 | not reconfirmed |
| Z.AI (Zhipu GLM) | GLM-4.6V-FlashX | $0.04 | official page | 2026-11-04 | 2026-11-04 |
| Groq | openai/gpt-oss-20b | $0.075 | official page | 2026-10-28 | 2026-11-04 |
| DeepInfra | Llama-3.3-70B-Instruct-Turbo | $0.1 | official page | 2026-10-28 | 2026-11-04 |
Full list: /cheapest/cheap-volume-llm
Questions
- What does Agent Depot pick for High-volume cheap LLM inference?
- Fireworks AI is the stored Depot Pick for the job cheap-volume-llm. pricePerf + reliability + dx. Open-model hosts win; Gemini Flash also excellent closed-model value.
- What is the runner-up?
- Together AI, then Groq. The ranking is editorial (scores as of 2026-10-28) and is not chosen by price.
- When should I not use Fireworks AI?
- Need absolute lowest latency on a supported model (Groq/Cerebras) or compliance requires Bedrock/Azure.
- How much does Fireworks AI cost?
- Embeddings up to 150M params: $0.008 per million input tokens, copied from https://fireworks.ai/pricing (price as of 2026-10-28). More rows are in the table on this page.
Picks are editorial and dated; community reports never change a pick or a score. Method: /methodology · Corrections: /corrections · Compare: Fireworks AI vs Together AI