Skip to content
inetGeek

LLM API pricing, compared

Each provider is profiled from its own documentation, with the source recorded against every figure. Token prices name the model they were read from, because “flagship” is a role rather than a product — and because these prices move, every figure here is re-verified monthly rather than on the site's usual four-month cycle.

Updated Sources last checked

LLM APIs

AnthropicHosted inference for the Claude model family, sold per million tokens with prompt caching and a batch tier.BasetenModel inference platform with per-token APIs for open models and dedicated deployments billed by the minute.DeepInfraServerless inference for open-weight models with published cached-input rates, plus dedicated GPUs billed by the minute.DeepSeekHosted inference for the DeepSeek model family, priced per million tokens with peak and off-peak rates.Fireworks AIServerless inference for open-weight models from other labs, priced per token, with fine-tuning and dedicated GPUs billed separately.Google Gemini APIHosted inference for the Gemini model family, with a free tier and per-million-token paid pricing.Novita AIServerless inference across 200+ open models with per-model cache-read rates, plus on-demand and spot GPU instances.OpenAIHosted inference for the GPT model family, sold per million tokens with separate short- and long-context rates.Together AIServerless per-token inference for open-weight models from other labs, plus dedicated GPU endpoints, hourly GPU clusters and fine-tuning, paid from prepaid credits.xAIHosted inference for the Grok model family from xAI (SpaceXAI in its docs), priced per million tokens with a higher rate once a prompt reaches 200k tokens.
ProviderCheapest model, per million input tokensContext window
Anthropic$1 per million input tokens for Claude Haiku 4.5, the cheapest current model.1M tokens on Claude Fable 5.1, Claude Opus 5 and Claude Sonnet 5; 200K on Claude Haiku 4.5.
Baseten$0.10 per million input tokens for GPT OSS 120B, with output at $0.50. GLM-5.3-Flash is $0.15 in and $0.50 out.1,048K tokens on the DeepSeek V4 line and the GLM 5.2/5.3 family, the widest it serves; 262K on the Kimi K2 models, 200K on GLM 4.7 and 128K on GPT OSS 120B. Baseten's table publishes these in thousands, so the magnitude is 1,048,000 rather than the 1,048,576 a power-of-two reading would give — the vendor's own figure, not a conversion of it.
DeepInfra$0.06 per million input tokens for DeepSeek-V4-Flash-0731, with cached input at $0.015 and output at $0.18. DeepSeek itself charges $0.44 at peak for the same checkpoint and Fireworks $0.22.1024k tokens on the DeepSeek V4 models it serves; 160k on the V3 generation. The window is the model's rather than a platform limit.
DeepSeek$0.3 per million input tokens for deepseek-flash at peak on a cache miss, $0.15 off-peak; output is $1.2 peak and $0.6 off-peak.1M tokens on both models in the current line, deepseek-flash and deepseek-v4-pro.
Fireworks AI$0.15 per million input tokens for GLM 5.3 Flash, with output at $0.50. DeepSeek-V4-Flash is $0.22 in and $0.66 out — the same model DeepSeek itself sells at $0.44 peak.1,048,576 tokens on Kimi K3, GLM 5.3 and both DeepSeek V4 variants; 262,144 on Qwen3.8 and 131,072 on Muse Glimmer 30B. The window is the model's, not Fireworks'.
Google Gemini API$0.30 per million input tokens for Gemini 3.5 Flash-Lite, with output at $2.50.1,048,576 tokens for Gemini 3.8 Flash. Read from the Vertex AI model card: the Gemini API docs build their model cards client-side and publish no figure a fetch can read.
Novita AI$0.075 per million input tokens for GLM 5.3 Flash at short context, rising to $0.15 above the threshold, with cache reads at $0.015 and output at $0.25. Fireworks serves the same model at $0.15.1M tokens on the DeepSeek V4 line and GLM 5.3, published as "1M" rather than an exact figure — so it is recorded at 1,000,000 and may in practice be the 1,048,576 other providers state. Qwen3.8 is listed at 977K.
OpenAI$0.10 per million input tokens for gpt-6-luna at short context.1,050,000 tokens on gpt-6-astra.
Together AI$0.09 per million input tokens for Qwen3.8 Flash, with output at $0.28 - the cheapest paid chat model on the pricing page. DeepSeek V4 Flash 0731 is $0.14 in and $0.28 out. Ternary Bonsai 27B is listed at $0.00 for both input and output.1,048,576 tokens on Kimi K3 and on DeepSeek V4 Pro 0813 and V4 Flash 0731; 1,048,575 on GLM-5.3 and GLM-5.2; 1,000,000 on Qwen3.8 Flash and the Qwen3.6/3.7 Plus models; 131,072 on gpt-oss-120b and Llama 3.3 70B. The window is the model's, as listed in the docs catalogue.
xAI$1.25 per million input tokens for grok-4.3 below 200k prompt tokens (output $2.50), the cheapest generally available text model. The early-access coding model grok-build-0.1 is listed at $1.00.500,000 tokens on grok-4.7; 1M on grok-4.3 and the grok-4.20 models, 256k on grok-build-0.1.

Which to consider

Anthropic

Teams that want the full 1M-token window at the standard per-token rate rather than a long-context surcharge, and that can use caching and batch to cut a frontier-model bill roughly in half.

Editorial · Palash Bagchi

Baseten

Teams that want per-token APIs and dedicated GPUs from one vendor with no plan fee — minute-granularity billing means a model that runs for ten minutes costs ten minutes.

Editorial · Palash Bagchi

DeepInfra

Running open-weight models at the lowest published per-token rates here — DeepSeek-V4-Flash at $0.06 per million input is a fraction of what the lab itself charges — with dedicated H100s at $2.20 an hour if throughput demands it.

Editorial · Palash Bagchi

DeepSeek

Batch-shaped work that can run outside weekday mornings UTC, where off-peak halves an already low rate and a cache hit costs about 3% of a cache miss.

Editorial · Palash Bagchi

Fireworks AI

Running open-weight models without operating GPUs — the same DeepSeek and Qwen checkpoints the labs publish, served per token, with a path to dedicated hardware if throughput demands it.

Editorial · Palash Bagchi

Google Gemini API

Getting to a working prototype without a card, then staying on the cheapest frontier-adjacent rates in this set — $0.75 per million input tokens is well under half what Anthropic and OpenAI charge for their top models.

Editorial · Palash Bagchi

Novita AI

Breadth — 200+ models under one API with a cache-read rate published next to most of them, and an H100 SXM at $3.39 an hour if the same vendor should run the dedicated side too.

Editorial · Palash Bagchi

OpenAI

Work that fits inside the short-context rate, where the cheapest tier is $0.20 per million input tokens and the batch tier halves everything above it.

Editorial · Palash Bagchi

Together AI

Open-weight models from many labs behind one prepaid API: Kimi, Qwen, DeepSeek, GLM and Gemma, from $0.09 per million input tokens, with dedicated GPU endpoints and H100 clusters at $3.99 an hour when serverless limits are not enough.

Editorial · Palash Bagchi

xAI

Grok models on pay-as-you-go credits: grok-4.7 at $2 in / $6 out per million tokens with a 500k context, or grok-4.3 at $1.25 / $2.50 with a 1M context and 20% off through the Batch API, plus a US-only endpoint when data must stay in the US.

Editorial · Palash Bagchi

LLM APIs in the news

Headlines from other publications, not checked by us · collected

Selected by a search on this sector, refreshed weekly. Nobody here has verified these claims — where something material is confirmed, it is written into the page above with its source, the way every other fact on this site is.

Comments

When these numbers change, hear about it. Sources are re-checked monthly; a repricing goes out as a short note.

Confirm by email; unsubscribe from any issue. Your address goes to Kit and nowhere else — what we do with it.

Every page here is sourced and dated.