BroadRouter

LLM API price comparisons

Updated

651 side-by-side API price comparisons of 183 models from 23 makers: GPT, Claude, Gemini, Grok, DeepSeek, Qwen, GLM, Kimi, Llama, Gemma and more. Each page shows both models' cheapest standard price, the ratio at 3 input tokens to 1 output, workload costs, the crossover mix and the providers that sell both. Price only, no quality claims.

Every price is read from the provider's own pricing page (41 providers, newest check 2026-11-15) and linked on each comparison. These pages compare price only; for quality, speed and accuracy, use independent benchmarks and your own tests. For BroadNet's own models side by side (features, plans and limits), see compare BroadNet models.

Jump to a maker:

How the pairs were picked

Of the 651 pairs, 45 pairs tie on the 3:1 blend, and in 287 pairs the cheaper model changes with the input:output mix. Each model below lists every comparison it is in, with the other model's cheapest 3:1 price relative to it.

Qwen model price comparisons

Qwen 2.5 7B USD 0.05 at 3:1 · 7 comparisons

Qwen-MT Plus USD 0.375 at 3:1 · 4 comparisons

Qwen2.5 72B USD 0.37 at 3:1 · 6 comparisons

Qwen2.5 Coder 32B USD 0.745 at 3:1 · 4 comparisons

Qwen3 30B A3B USD 0.122 at 3:1 · 8 comparisons

Qwen3 32B USD 0.13 at 3:1 · 7 comparisons

Qwen3 235B A22B USD 0.40 at 3:1 · 5 comparisons

Qwen3 235B A22B 2507 USD 0.1531 at 3:1 · 5 comparisons

Qwen3 235B A22B Thinking 2507 USD 1.212 at 3:1 · 5 comparisons

Qwen3 Coder 480B A35B USD 0.4375 at 3:1 · 5 comparisons

Qwen3 Coder Next USD 0.29 at 3:1 · 5 comparisons

Qwen3 Next 80B A3B Thinking USD 0.4125 at 3:1 · 5 comparisons

Qwen3 VL 30B A3B Instruct USD 0.2375 at 3:1 · 6 comparisons

Qwen3 VL 235B A22B USD 0.37 at 3:1 · 4 comparisons

Qwen3-8B USD 0.06 at 3:1 · 7 comparisons

Qwen3-14B USD 0.1225 at 3:1 · 6 comparisons

Qwen3-30B-A3B-2507 USD 0.1425 at 3:1 · 6 comparisons

Qwen3-VL 8B USD 0.305 at 3:1 · 5 comparisons

Qwen3.5 9B USD 0.1125 at 3:1 · 6 comparisons

Qwen3.5 27B USD 0.5875 at 3:1 · 10 comparisons

Qwen3.5 35B A3B USD 0.355 at 3:1 · 10 comparisons

Qwen3.5 122B A10B USD 0.715 at 3:1 · 7 comparisons

Qwen3.5 397B A17B USD 1.087 at 3:1 · 6 comparisons

Qwen3.6 27B USD 0.6875 at 3:1 · 8 comparisons

Qwen3.6 35B A3B USD 0.2975 at 3:1 · 9 comparisons

Qwen3.6 Plus USD 1.125 at 3:1 · 6 comparisons

Qwen3.7 Max USD 1.875 at 3:1 · 6 comparisons

Qwen3.7 Plus USD 0.56 at 3:1 · 6 comparisons

Qwen3.8 2.4T A95B USD 1.75 at 3:1 · 10 comparisons

Qwen3.8 27B USD 0.3417 at 3:1 · 15 comparisons

Qwen3.8 Flash USD 0.1803 at 3:1 · 10 comparisons

Qwen3.8 Max USD 2.475 at 3:1 · 9 comparisons

Qwen3.8 Max (0902) USD 2.475 at 3:1 · 1 comparison

OpenAI model price comparisons

GPT OSS Safeguard 20B USD 0.1025 at 3:1 · 4 comparisons

GPT-4.1 USD 3.50 at 3:1 · 8 comparisons

GPT-4.1 Mini USD 0.70 at 3:1 · 8 comparisons

GPT-4o USD 4.375 at 3:1 · 4 comparisons

GPT-4o mini USD 0.2625 at 3:1 · 4 comparisons

GPT-5 USD 3.438 at 3:1 · 10 comparisons

GPT-5 Mini USD 0.6875 at 3:1 · 11 comparisons

GPT-5 Nano USD 0.1375 at 3:1 · 8 comparisons

GPT-5 Pro USD 41.25 at 3:1 · 7 comparisons

GPT-5.2 USD 4.813 at 3:1 · 6 comparisons

GPT-5.2 Pro USD 57.75 at 3:1 · 6 comparisons

GPT-5.4 USD 5.625 at 3:1 · 6 comparisons

GPT-5.4 Mini USD 1.688 at 3:1 · 6 comparisons

GPT-5.4 Nano USD 0.4625 at 3:1 · 7 comparisons

GPT-5.4 Pro USD 67.50 at 3:1 · 6 comparisons

GPT-5.5 USD 11.25 at 3:1 · 10 comparisons

GPT-5.6 Luna USD 0.45 at 3:1 · 9 comparisons

GPT-5.6 Sol USD 8 at 3:1 · 10 comparisons

GPT-5.6 Terra USD 4.50 at 3:1 · 9 comparisons

GPT-6 Astra USD 20 at 3:1 · 8 comparisons

GPT-6 Luna USD 0.20 at 3:1 · 13 comparisons

GPT-6 Sol USD 4 at 3:1 · 9 comparisons

GPT-6.1 Sol USD 4 at 3:1 · 11 comparisons

GPT-6.1 Terra USD 2 at 3:1 · 7 comparisons

gpt-oss-20b USD 0.036 at 3:1 · 7 comparisons

gpt-oss-120b USD 0.065 at 3:1 · 9 comparisons

DeepSeek model price comparisons

DeepSeek R1 USD 0.92 at 3:1 · 9 comparisons

DeepSeek R1 0528 USD 0.695 at 3:1 · 10 comparisons

DeepSeek R1 Distill Llama 8B USD 0.2025 at 3:1 · 5 comparisons

DeepSeek R1 Distill Llama 70B USD 0.375 at 3:1 · 6 comparisons

DeepSeek R1 Distill Qwen 7B USD 0.125 at 3:1 · 5 comparisons

DeepSeek R1 Distill Qwen 14B USD 0.20 at 3:1 · 4 comparisons

DeepSeek R1 Distill Qwen 32B USD 0.60 at 3:1 · 7 comparisons

DeepSeek V3 USD 0.4375 at 3:1 · 7 comparisons

DeepSeek V3 0324 USD 0.5025 at 3:1 · 5 comparisons

DeepSeek V3.1 USD 0.425 at 3:1 · 7 comparisons

DeepSeek V3.2 USD 0.234 at 3:1 · 11 comparisons

DeepSeek V4 Flash USD 0.1125 at 3:1 · 11 comparisons

DeepSeek V4 Pro USD 1.196 at 3:1 · 12 comparisons

DeepSeek V4 Pro 0813 USD 1.30 at 3:1 · 6 comparisons

DeepSeek V4.1 Flash USD 0.21 at 3:1 · 18 comparisons

DeepSeek V4.1 Flash 1017 USD 0.147 at 3:1 · 7 comparisons

DeepSeek-V3.2-Exp USD 0.305 at 3:1 · 5 comparisons

DeepSeek-V4-Flash-Vision-Exp USD 0.3234 at 3:1 · 5 comparisons

Google model price comparisons

Gemini 2.5 Flash USD 0.85 at 3:1 · 2 comparisons

Gemini 2.5 Flash-Lite USD 0.175 at 3:1 · 6 comparisons

Gemini 2.5 Pro USD 3.438 at 3:1 · 7 comparisons

Gemini 3.1 Pro Preview USD 4.50 at 3:1 · 11 comparisons

Gemini 3.5 Flash USD 3.375 at 3:1 · 5 comparisons

Gemini 3.5 Flash-Lite USD 0.85 at 3:1 · 7 comparisons

Gemini 3.6 Flash USD 1.50 at 3:1 · 2 comparisons

Gemini 3.7 Flash USD 1.50 at 3:1 · 2 comparisons

Gemini 3.8 Flash USD 1.50 at 3:1 · 7 comparisons

Gemini 4 Argon USD 4 at 3:1 · 11 comparisons

Gemini 4 Hydrogen USD 3 at 3:1 · 6 comparisons

Gemma 3 4B USD 0.05 at 3:1 · 7 comparisons

Gemma 3 12B USD 0.075 at 3:1 · 6 comparisons

Gemma 3 27B USD 0.10 at 3:1 · 6 comparisons

Gemma 4 26B A4B USD 0.1375 at 3:1 · 13 comparisons

Gemma 4 31B USD 0.16 at 3:1 · 9 comparisons

gemma-4-31B-it-turbo USD 0.1525 at 3:1 · 5 comparisons

Z.ai model price comparisons

GLM 4.5 Air USD 0.31 at 3:1 · 6 comparisons

GLM 4.5V USD 0.90 at 3:1 · 7 comparisons

GLM 4.6 USD 0.76 at 3:1 · 6 comparisons

GLM 4.6V USD 0.45 at 3:1 · 6 comparisons

GLM 4.7 USD 0.7375 at 3:1 · 9 comparisons

GLM 4.7 Flash USD 0.145 at 3:1 · 9 comparisons

GLM 5 USD 0.93 at 3:1 · 7 comparisons

GLM 5.1 USD 1.40 at 3:1 · 6 comparisons

GLM 5.2 USD 0.8719 at 3:1 · 7 comparisons

GLM 5.3 USD 1.047 at 3:1 · 17 comparisons

GLM 5.3 Flash USD 0.1187 at 3:1 · 10 comparisons

GLM 5.3 FlashX USD 0.59 at 3:1 · 10 comparisons

GLM-4.5-FP8 USD 1 at 3:1 · 7 comparisons

Meta model price comparisons

Llama 3 8B USD 0.09312 at 3:1 · 8 comparisons

Llama 3 70B USD 2.862 at 3:1 · 11 comparisons

Llama 3.1 8B Instruct USD 0.025 at 3:1 · 6 comparisons

Llama 3.1 70B USD 0.72 at 3:1 · 10 comparisons

Llama 3.2 1B USD 0.0705 at 3:1 · 5 comparisons

Llama 3.2 3B USD 0.1219 at 3:1 · 8 comparisons

Llama 3.2 11B Vision USD 0.16 at 3:1 · 4 comparisons

Llama 3.3 70B USD 0.155 at 3:1 · 9 comparisons

Llama 4 Maverick USD 0.3875 at 3:1 · 10 comparisons

Llama 4 Scout USD 0.15 at 3:1 · 7 comparisons

Muse Glimmer 30B USD 0.50 at 3:1 · 4 comparisons

Muse Spark Open USD 2 at 3:1 · 4 comparisons

Anthropic model price comparisons

Claude Fable 5 USD 20 at 3:1 · 8 comparisons

Claude Fable 5.1 USD 20 at 3:1 · 2 comparisons

Claude Fable 5.2 USD 16 at 3:1 · 7 comparisons

Claude Haiku 4.5 USD 2 at 3:1 · 6 comparisons

Claude Haiku 5.5 USD 0.20 at 3:1 · 11 comparisons

Claude Opus 5 USD 10 at 3:1 · 12 comparisons

Claude Opus 5.5 USD 8 at 3:1 · 13 comparisons

Claude Sonnet 4.6 USD 6 at 3:1 · 8 comparisons

Claude Sonnet 5 USD 4 at 3:1 · 3 comparisons

Claude Sonnet 5.5 USD 4 at 3:1 · 17 comparisons

Moonshot AI model price comparisons

Kimi K2 USD 1.002 at 3:1 · 8 comparisons

Kimi K2 0905 USD 1.075 at 3:1 · 7 comparisons

Kimi K2 Thinking USD 0.90 at 3:1 · 6 comparisons

Kimi K2.5 USD 0.90 at 3:1 · 7 comparisons

Kimi K2.6 USD 0.7425 at 3:1 · 9 comparisons

Kimi K2.7 Code USD 1.20 at 3:1 · 5 comparisons

Kimi K3 USD 3.763 at 3:1 · 18 comparisons

Kimi K3.1 Pro USD 3.80 at 3:1 · 13 comparisons

Mistral AI model price comparisons

Ministral 3 3B USD 0.10 at 3:1 · 5 comparisons

Ministral 3 8B USD 0.15 at 3:1 · 7 comparisons

Ministral 3 14B USD 0.20 at 3:1 · 7 comparisons

Mistral Large 3 USD 0.75 at 3:1 · 6 comparisons

Mistral Medium 3 USD 0.80 at 3:1 · 4 comparisons

Mistral Small 3.1 24B USD 0.15 at 3:1 · 6 comparisons

Mistral Small 3.2 USD 0.1062 at 3:1 · 5 comparisons

Mistral-Nemo-2407 USD 0.02925 at 3:1 · 4 comparisons

Mistral-Small-24B-2501 USD 0.0575 at 3:1 · 5 comparisons

MiniMax model price comparisons

MiniMax M2 USD 0.4856 at 3:1 · 6 comparisons

Minimax M2.1 USD 0.525 at 3:1 · 6 comparisons

MiniMax M2.5 USD 0.44 at 3:1 · 8 comparisons

MiniMax M2.7 USD 0.3675 at 3:1 · 8 comparisons

MiniMax M3 USD 0.4125 at 3:1 · 15 comparisons

MiniMax M3.1 Flash USD 0.2625 at 3:1 · 7 comparisons

NVIDIA model price comparisons

Nemotron 3 Nano 30B A3B USD 0.0875 at 3:1 · 7 comparisons

Nemotron 3 Super USD 0.1563 at 3:1 · 6 comparisons

Nemotron 3 Ultra USD 0.9125 at 3:1 · 7 comparisons

Nemotron 3.5 Lightning 30B A3B USD 0.085 at 3:1 · 7 comparisons

Nemotron 3.5 Super VL USD 0.175 at 3:1 · 5 comparisons

Tencent model price comparisons

Hy3 USD 0.231 at 3:1 · 7 comparisons

Hy4 USD 0.6254 at 3:1 · 9 comparisons

Hy4 Preview USD 1.251 at 3:1 · 5 comparisons

xAI model price comparisons

Grok 4.3 USD 1.563 at 3:1 · 6 comparisons

Grok 4.5 USD 3 at 3:1 · 2 comparisons

Grok 4.6 USD 3 at 3:1 · 1 comparison

Grok 4.7 USD 3 at 3:1 · 9 comparisons

Xiaomi model price comparisons

MiMo v2.5 USD 0.175 at 3:1 · 6 comparisons

MiMo V2.6 Flash USD 0.14 at 3:1 · 6 comparisons

MiMo V2.6 Pro USD 0.4078 at 3:1 · 5 comparisons

StepFun model price comparisons

Step 3.7 Flash USD 0.4375 at 3:1 · 5 comparisons

Step 5 Preview USD 1.425 at 3:1 · 4 comparisons

Step-3.5-Flash USD 0.15 at 3:1 · 5 comparisons

Sao10K model price comparisons

L3.1-70B-Euryale-v2.2 USD 0.72 at 3:1 · 6 comparisons

Llama 3 8B Stheno V3.2 USD 0.05 at 3:1 · 6 comparisons

Cohere model price comparisons

Command A+ USD 1.40 at 3:1 · 4 comparisons

Command R USD 0.2625 at 3:1 · 7 comparisons

IBM model price comparisons

Granite 4.2 8B USD 0.0825 at 3:1 · 6 comparisons

Ricursive AI model price comparisons

Ripple-65B Preview USD 0.30 at 3:1 · 6 comparisons

inclusionAI model price comparisons

Ling 3.0 Flash USD 0.09 at 3:1 · 5 comparisons

Microsoft model price comparisons

Phi-4 USD 0.0875 at 3:1 · 4 comparisons

Reflection AI model price comparisons

Beam 1.0 501B A23B USD 0.60 at 3:1 · 4 comparisons

Thinking Machines model price comparisons

Inkling 2 Small USD 0.4512 at 3:1 · 4 comparisons

How were the pairs picked?

Both models of every pair need an indexed price page: at least 2 providers with a standard price read from their own pricing page, one of them paid (the maker's own API counts as one). A pair that later falls below that rule is kept out of search results, not deleted. Beyond that, the pairs come from fixed rules over the 198 indexed models, using only the maker, the model's name and the sourced prices:

  • Hand-picked (67): the original list of rivals and releases developers weigh against each other.
  • Successors (59): each release and the previous release in the same line, by version in the name (Claude Opus, GPT Sol, Gemini Flash, GLM, Kimi, MiniMax ...). Sized open-weight lines (Qwen, Gemma, Llama, Mistral Small, Nemotron) pair a model with the newest earlier generation of about the same size, and dated snapshots with their base model.
  • Same family (61): neighbouring tiers of one generation (Luna, Terra, Sol; Haiku, Sonnet, Opus; Flash-Lite, Flash, Pro), neighbouring sizes of one generation, and variants (thinking, code, vision, Pro) against the main model.
  • Same-tier rivals (422): models are bucketed by price band, the quarter of the indexed models their cheapest 3:1 price falls in (below USD 0.1591, USD 0.1591 to USD 0.4569, USD 0.4569 to USD 1.40, above USD 1.40 per 1M tokens), and size class (a tier word such as Flash, Mini or Haiku, or Pro, Opus or Max in the name; else the parameter count: 40B or less small, 200B or more large). Each model is paired with the 4 closest-priced models from other makers in its bucket, topped up from its price band when the bucket has fewer than 3.
  • BroadNet's chat models (16): each of BroadNet's four chat models against the 6 models from other makers in its price band that the most providers sell.
  • Open-weight substitutes (26): each closed model (Claude, GPT except gpt-oss, the o-series, Gemini, Grok) against the 2 closest-priced open-weight models in its price band, at most 8× apart, each open model used at most 6 times.
  • No near-duplicates: when two models from one maker have identical standard prices at every provider that sells both, only one takes part in the rules. When two comparisons would read the same apart from version numbers (for example Claude Opus 5 and Opus 5.5 against the same model), both pages stay, but only one is offered to search engines.

All models: API prices compared · Cheapest LLM APIs right now · LLM API cost calculator · Download all prices (CSV, JSON)

Read as MarkdownAsk ChatGPTAsk ClaudeAsk Perplexity