API Pricing Comparison
Compare AI model API prices across different providers and channels. Find the cheapest option for your use case.
智谱 AI (ChatGLM) • 1,048,576 tokens • Released Jun 17, 2026 • Compare plans
| Channel | Region | Input / 1M | Output / 1M | Savings | Visit |
|---|---|---|---|---|---|
智谱 AI (ChatGLM)🇨🇳 | Official | ¥8.00 | ¥28.00 | Official baseline | |
Fireworks AI🌍 | Aggregator | $1.40 | $4.40 | - | |
Together AI🌍 | Aggregator | $1.40 | $4.40 | - | |
XycAi🇨🇳 | Reseller | ¥5.20 | ¥18.20 | -35% | |
OpenRouter29 variants🌍 | Aggregator | $0.65 | $2.04 | -44% | |
Show all price variantsOpenRouter Auto / default routeDefault $0.65 / $2.04 per 1M default_route Baidu | z-ai/glm-5.2-20260616 $0.55 / $1.74 per 1M upstream_provider StreamLake | z-ai/glm-5.2-20260616 $0.56 / $1.75 per 1M upstream_provider DeepInfra | z-ai/glm-5.2-20260616 $0.56 / $1.80 per 1M upstream_provider Ambient | z-ai/glm-5.2-20260616 $0.60 / $2.00 per 1M upstream_provider Baidu | z-ai/glm-5.2-20260616 $0.65 / $2.27 per 1M upstream_provider Novita | z-ai/glm-5.2-20260616 $0.65 / $2.04 per 1M upstream_provider DigitalOcean | z-ai/glm-5.2-20260616 $0.70 / $2.20 per 1M upstream_provider CoreWeave | z-ai/glm-5.2-20260616 $0.76 / $2.42 per 1M upstream_provider AtlasCloud | z-ai/glm-5.2-20260616 $0.94 / $2.95 per 1M upstream_provider Alibaba | z-ai/glm-5.2-20260616 $0.97 / $3.04 per 1M upstream_provider SiliconFlow | z-ai/glm-5.2-20260616 $1.19 / $3.74 per 1M upstream_provider Inceptron | z-ai/glm-5.2-20260616 $1.25 / $3.34 per 1M upstream_provider Phala | z-ai/glm-5.2-20260616 $1.26 / $3.00 per 1M upstream_provider Venice | z-ai/glm-5.2-20260616 $1.40 / $4.40 per 1M upstream_provider BaseTen | z-ai/glm-5.2-20260616 $1.40 / $4.40 per 1M upstream_provider Wafer | z-ai/glm-5.2-20260616 $1.40 / $4.40 per 1M upstream_provider Cloudflare | z-ai/glm-5.2-20260616 $1.40 / $4.40 per 1M upstream_provider Fireworks | z-ai/glm-5.2-20260616 $1.40 / $4.40 per 1M upstream_provider Parasail | z-ai/glm-5.2-20260616 $1.40 / $4.40 per 1M upstream_provider GMICloud | z-ai/glm-5.2-20260616 $1.40 / $4.40 per 1M upstream_provider Together | z-ai/glm-5.2-20260616 $1.40 / $4.40 per 1M upstream_provider Friendli | z-ai/glm-5.2-20260616 $1.40 / $4.40 per 1M upstream_provider Z.AI | z-ai/glm-5.2-20260616 $1.40 / $4.40 per 1M upstream_provider BaseTen | z-ai/glm-5.2-20260616 $2.10 / $6.60 per 1M upstream_provider Fireworks | z-ai/glm-5.2-20260616 $2.10 / $6.60 per 1M upstream_provider Fireworks | z-ai/glm-5.2-20260616 $2.10 / $6.60 per 1M upstream_provider Decart | z-ai/glm-5.2-20260616 $2.25 / $8.00 per 1M upstream_provider Alibaba | z-ai/glm-5.2-20260616 $2.31 / $7.26 per 1M upstream_provider | |||||
智谱 AI (ChatGLM) • 128,000 tokens • Released Jun 17, 2026 • Compare plans
| Channel | Region | Input / 1M | Output / 1M | Savings | Visit |
|---|---|---|---|---|---|
OpenRouter2 variants🌍 | Aggregator | $0.00 | $0.00 | - | |
Show all price variantsOpenRouter Auto / default routeDefault $0.00 / $0.00 per 1M default_route Decart | z-ai/glm-5.2-20260616:free $0.00 / $0.00 per 1M upstream_provider | |||||
智谱 AI (ChatGLM) • 1,048,576 tokens • Released Jun 17, 2026 • Compare plans
| Channel | Region | Input / 1M | Output / 1M | Savings | Visit |
|---|---|---|---|---|---|
OpenRouter2 variants🌍 | Aggregator | $0.70 | $2.20 | - | |
Show all price variantsOpenRouter Auto / default routeDefault $0.70 / $2.20 per 1M default_route Fireworks | z-ai/glm-5.2-20260616:batch $0.70 / $2.20 per 1M upstream_provider | |||||
智谱 AI (ChatGLM) • N/A • Added Aug 18, 2026 • Compare plans
| Channel | Region | Input / 1M | Output / 1M | Savings | Visit |
|---|---|---|---|---|---|
Fireworks AI🌍 | Aggregator | $2.10 | $6.60 | - |
智谱 AI (ChatGLM) • N/A • Added Aug 18, 2026 • Compare plans
| Channel | Region | Input / 1M | Output / 1M | Savings | Visit |
|---|---|---|---|---|---|
Fireworks AI🌍 | Aggregator | $2.10 | $6.60 | - |
智谱 AI (ChatGLM) • N/A • Added Aug 18, 2026 • Compare plans
| Channel | Region | Input / 1M | Output / 1M | Savings | Visit |
|---|---|---|---|---|---|
Mistral AI🌍 | Official | $1.40 | $4.40 | Official baseline |
Frequently asked questions
Which model has the cheapest API?
Of the 348 models and 629 channels tracked here, the cheapest in USD-normalised terms is gpt-5.6-luna via XiuRouter at $0.01 per 1M input tokens. Prices are normalised to USD for comparison after being read in each channel's own currency; confirm on the provider's page.
How are API prices in different currencies compared?
Every channel is quoted in its own currency (USD or CNY), then normalised to USD at the current exchange rate for comparison — so ¥3 is never compared as if it were $3. Each table row still shows the channel's original currency and price. We track 629 channels across 23 providers.
What is the difference between official, cloud, aggregator and reseller channels?
Official channels are the model producer's own API (OpenAI, Anthropic, DeepSeek); cloud channels host the same model on Azure, AWS Bedrock or Google Vertex; aggregators (OpenRouter, SiliconFlow) unify many models under one account and bill; resellers resell capacity. They differ in price, rate limits, compliance and who invoices you.
Can I call these APIs directly from mainland China?
Each channel is flagged for direct mainland-China access. About 169 models have at least one directly reachable channel (including domestic producers and some aggregators); international official channels normally require a proxy or VPN. Use the "China access only" filter on the page.
Besides token-priced APIs, are there subscription plans?
Yes. Alongside metered API pricing we compare subscriptions such as ChatGPT Plus/Pro/Team, Claude Pro/Max/Team, Gemini Advanced, Kimi and GLM, with monthly and annual pricing. Open "Compare plans" to see every plan that includes a given model.
How often are these prices updated, and how reliable is the data?
API prices are scraped daily from official pricing pages and every tracked channel, with significant changes recorded in a price-history log. A daily read-only data audit cross-checks each channel against the model producer's published price and flags stale or inconsistent rows for correction.