Models & rate card
Prices are USD per 1M tokens. The struck-through number is the cloud vendor's official list price; the highlighted number is your discounted price.
23 / 23
| Model | Served on | Capabilities | Context | Input | Output | Cached input | Health · 24h |
|---|---|---|---|---|---|---|---|
Gemini 3.1 Pro×0.6 gemini-3.1-pro | Google Google Cloud · Vertex AI | 1.0M | $1.20$2.00 | $7.20$12.00 | $0.12$0.2 | 100% | |
Gemini 3.8 Flash×0.6 gemini-3.8-flash | Google Google Cloud · Vertex AI | 1.0M | $0.45$0.75 | $2.25$3.75 | $0.045$0.075 | 100% | |
Gemini 3.5 Flash×0.6 gemini-3.5-flash | Google Google Cloud · Vertex AI | 1.0M | $0.9$1.50 | $5.40$9.00 | $0.09$0.15 | 100% | |
Gemini 3.5 Flash-Lite×0.6 gemini-3.5-flash-lite | Google Google Cloud · Vertex AI | 1.0M | $0.18$0.3 | $1.50$2.50 | $0.018$0.03 | 100% | |
Gemini 3.1 Flash-Lite×0.6 gemini-3.1-flash-lite | Google Google Cloud · Vertex AI | 1.0M | $0.15$0.25 | $0.9$1.50 | $0.015$0.025 | 100% | |
Llama 4 Maverick×0.6 llama-4-maverick | Meta Google Cloud · Vertex AI | 1.0M | $0.21$0.35 | $0.69$1.15 | — | 100% | |
Qwen3 235B A22B×0.6 qwen3-235b | Alibaba Google Cloud · Vertex AI | 262K | $0.132$0.22 | $0.528$0.88 | — | 100% | |
GLM-4.7×0.6 glm-4.7 | Zhipu Google Cloud · Vertex AI | 200K | $0.36$0.6 | $1.32$2.20 | — | 100% | |
Kimi K2 Thinking×0.6 kimi-k2-thinking | Moonshot Google Cloud · Vertex AI | 262K | $0.36$0.6 | $1.50$2.50 | — | 100% | |
Claude Opus 5×0.6 claude-opus-5 | Anthropic AWS · Amazon Bedrock | 1M | $3.00$5.00 | $15.00$25.00 | $0.3$0.5 | 100% | |
Claude Sonnet 5×0.6 claude-sonnet-5 | Anthropic AWS · Amazon Bedrock | 1M | $1.20$2.00 | $6.00$10.00 | $0.12$0.2 | 100% | |
Claude Opus 4.8×0.6 claude-opus-4.8 | Anthropic AWS · Amazon Bedrock | 1M | $3.00$5.00 | $15.00$25.00 | $0.3$0.5 | 100% | |
Claude Sonnet 4.6×0.6 claude-sonnet-4.6 | Anthropic AWS · Amazon Bedrock | 1M | $1.80$3.00 | $9.00$15.00 | $0.18$0.3 | 100% | |
Claude Haiku 4.5×0.6 claude-haiku-4.5 | Anthropic AWS · Amazon Bedrock | 200K | $0.6$1.00 | $3.00$5.00 | $0.06$0.1 | 100% | |
Gemma 4 26B A4B×0.6 gemma-4-26b | Google AWS · Amazon Bedrock | 131K | $0.078$0.13 | $0.24$0.4 | — | 100% | |
GPT-OSS 120B×0.6 gpt-oss-120b | OpenAI AWS · Amazon Bedrock | 131K | $0.09$0.15 | $0.36$0.6 | — | 100% | |
GPT-OSS 20B×0.6 gpt-oss-20b | OpenAI AWS · Amazon Bedrock | 131K | $0.042$0.07 | $0.09$0.15 | — | 100% | |
GLM-5×0.6 glm-5 | Zhipu AWS · Amazon Bedrock | 203K | $0.6$1.00 | $1.92$3.20 | — | 100% | |
Kimi K2.5×0.6 kimi-k2.5 | Moonshot AWS · Amazon Bedrock | 262K | $0.36$0.6 | $1.80$3.00 | — | 100% | |
Amazon Nova Premier×0.6 nova-premier | Amazon AWS · Amazon Bedrock | 1M | $1.50$2.50 | $7.50$12.50 | $0.375$0.625 | 100% | |
Amazon Nova Pro×0.6 nova-pro | Amazon AWS · Amazon Bedrock | 300K | $0.48$0.8 | $1.92$3.20 | — | 100% | |
Amazon Nova 2 Lite×0.6 nova-2-lite | Amazon AWS · Amazon Bedrock | 1M | $0.18$0.3 | $1.50$2.50 | — | 100% | |
Amazon Nova Lite×0.6 nova-lite | Amazon AWS · Amazon Bedrock | 300K | $0.036$0.06 | $0.144$0.24 | — | 100% |
Health is measured from real customer traffic: each bar is one hour, green ≥ 99% success, amber ≥ 90%, red below. Hours without requests are shown as healthy.
How metering works
Usage is metered per request. Failed requests (non-2xx) are never charged.
Each model lists the vendor's official price and your discounted price side by side; the multiplier next to the model name is what applies to it.
Every response carries the standard OpenAI usage object; the dashboard shows the cost of every request.