Scaleway Generative APIs Free Tier (1 million tokens + 60 minutes of speech transcription)

Generative APIs (Serverless) free tier: up to 1 million tokens free for models billed by tokens and up to 60 free minutes for transcription models billed by audio minutes, with pay-as-you-go billing after that

Official FAQ verbatim: "The Free Tier allows you to process, without incurring any costs, up to: 1,000,000 tokens for models billed by tokens; 60 minutes of audio transcription for models billed by audio minutes. After reaching this limit, you will be charged per million tokens processed and per minutes of audio processed." The pricing page likewise says "You benefit from a free tier on the first 1,000,000 tokens". The free allowance is applied first to the tokens with the highest unit price; when shared across several projects it is apportioned by each project's actual consumption. The API is compatible with the OpenAI SDK, and the official docs provide integration tutorials for OpenCode, Claude Code, Qwen Code, and LiteLLM. The official "Supported models" page lists DeepSeek V4 Flash 0731, GLM 5.2, Qwen3.5 / Qwen3-235B, gpt-oss, Mistral, Gemma, Llama, Whisper, and more; which specific models the free allowance applies to is as stated on that page.

⚠️The official text only says "up to 1 million tokens processed free" and does not state whether it is a one-time allowance or refreshes monthly (the billing example is shown for "the current month"), so go by the console before claiming. The official page states that all services are hosted in Europe. The sign-up flow was not completed, so whether the account forces a payment method / identity verification is as per the official source, which is why the threshold conservatively lists a credit card; reachability from mainland China was not tested. The supported-models page also lists models that have reached end of life (EOL), so do not call by old model names.