# packet.ai's Token Factory offers free API access to DeepSeek V4.1 Flash and other models for a limited time



packet.ai

Active · No claim needed

packet.ai's Token Factory is an OpenAI-compatible inference API for open-weight models. Sign up at dash.packet.ai with email or Google and complete one SMS phone check, and new and existing accounts can call DeepSeek-V4.1-Flash-NVFP4 \(131K context\), MiniMax-M3.1-Flash \(131K context\) and Qwen2.5-1.5B-Instruct \(33K context\) for free, at $0 for input and output tokens, with no credit card. Kimi K3 is only listed as coming soon. No free quota has been published; your account's rate and token limits only appear in the console under Billing → Usage → Limits. After the free period, usage is billed per token. Note: Packet's terms bar customers located in, resident in, organized in or headquartered in mainland China, Hong Kong, Macau and other restricted jurisdictions.

- Effective: Nov 17, 2026
- Ends: Dec 16, 2026 08:00 · Beijing Time (UTC+8)

## How to claim

Sign up at dash.packet.ai with your email or Google account, complete the SMS phone check, create a Token Factory key in the console, and set your SDK base_url to https://api.packet.ai/v1. Until the phone check is done, API calls are refused with a message that the team is not permitted to use Token Factory.

## Sources

- Source archive: https://packet.ai/token-factory-free
- Source archive: https://dash.packet.ai/docs/token-factory-billing
- Source archive: https://dash.packet.ai/docs/token-factory-models
- Source archive: https://dash.packet.ai/docs/token-factory-errors
- Source archive: https://packet.ai/blog/token-factory-free-llm-api
- Source archive: https://dash.packet.ai/terms
- Source excerpt: https://x.com/packet_ai/status/2122692289030140334
- Source archive: https://packet.ai/token-factory

Canonical entry: https://www.ailyre.com/source/?id=packet-token-factory-free-api-2026-11-17&lang=en
