synthetic

    Model pricing

    Models are charged per token. Usage-based requests pay the usage-based price; users on a subscription plan have their rate limit deducted at the subscription discounted rate.

    Usage-based priceSubscription discounted rate
    ModelInput / mtokInput (cached) / mtokOutput / mtok
    deepseek-ai/DeepSeek-V4.1-Flash-1017
    $0.60$0.06
    $0.03$0.003
    $1.20$0.30
    moonshotai/Kimi-K3.1-Pro
    $2.00$1.00
    $0.20$0.10
    $10.00$7.00
    openai/gpt-oss-120b
    $0.10$0.01
    $0.02$0.003
    $0.10$0.05
    Qwen/Qwen3.8-27B
    $0.45$0.23
    $0.09$0.05
    $2.20$1.10
    zai-org/GLM-4.7-Flash
    $0.10$0.01
    $0.02$0.003
    $0.50$0.05
    zai-org/GLM-5.3
    Beta
    $1.40$0.40
    $0.26$0.07
    $4.40$2.20
    zai-org/GLM-5.3-Flash
    $0.15$0.10
    $0.04$0.02
    $0.50$0.25