Fire Pass alternative: OpenScale Unlimited
Last updated: Nov 4, 2026
The short answer
Fire Pass is an experimental subscription from Fireworks AI that covers selected open model routes for personal agentic coding. OpenScale Unlimited is a flat monthly plan for your coding agents: unlimited tokens on DeepSeek V4.1 Flash and GLM 5.3 Flash, with no 5-hour window and no weekly limit. Plans start at $29 a month, and the first week is $5.
Side by side
| Fire Pass | OpenScale Unlimited | |
|---|---|---|
| Availability | Not open to new subscriptions | Open to everyone |
| Models | Only its enabled routes, today one Kimi K3 Fast route | DeepSeek V4.1 Flash and GLM 5.3 Flash |
| Allowed use | Personal, non-production agentic coding | Your own and your team's coding agents, while you build |
| Where it works | Not with the Responses API that Codex uses | Any tool that speaks the OpenAI or Anthropic API, Codex included |
| Usage limits | Covered while active, subject to its limits and terms | No token cap, no 5-hour window and no weekly limit on the included models |
Fireworks AI’s side is what its own pages say, checked Nov 4, 2026 (sources below). Anything we could not confirm there is left out.
Which one fits
Fire Pass is not open to new subscriptions, so a new user cannot start one today. Existing subscribers can still manage theirs.
OpenScale Unlimited is open to anyone, works with the Responses API that Codex uses, and covers DeepSeek and GLM models rather than Kimi.
OpenScale Unlimited plans
- Lite: 1 agent at once, $29 a month, first week $5
- Plus: 2 agents at once, $79 a month, first week $13
- Pro: 4 agents at once, $159 a month, first week $27
- Crew: 6 agents at once, $239 a month, first week $40
- Fleet: 10 agents at once, $399 a month, first week $67
Every plan has unlimited tokens on DeepSeek V4.1 Flash and GLM 5.3 Flash. Extra agent slots on Plus and larger plans are best effort: they open when there is room and are not guaranteed. See the plans.
Questions
- Is there a token limit on OpenScale Unlimited?
- No. DeepSeek V4.1 Flash and GLM 5.3 Flash have no token cap and no meter, all month. Other models and fast mode are pay as you go from your balance.
- What is the difference between the plans?
- Only how many agents run at once, from 1 on Lite to 10 on Fleet.
- What are extra agent slots?
- On Plus and larger plans, agents beyond your plan can run in extra agent slots. They are best effort: they open when there is room and are not guaranteed.
- Which coding agents work with it?
- Any agent that speaks the OpenAI Chat Completions or Responses API or the Anthropic Messages API, such as Claude Code, Codex, OpenCode, Cline and Kilo Code. A plan comes with its own API keys.
- Can I use a plan for an app that serves users?
- No. Plans are for your own and your team’s coding agents while you build. Apps, batch jobs and reselling fit pay as you go, billed per token.
Sources
- Fireworks AI Fire Pass docs, checked Nov 4, 2026
Other price comparisons are on the compare page.