Try Fusion with $5 in free credits

FUMO / PRICING

Pay only for what you use.

No monthly subscription and no seat fee. Add funds to your account, use any available Fusion model through one API, and pay for the tokens your requests consume.

Add balanceRead the API guide
Commitment
None
Billing cycle
Continuous balance deduction
Internal collaboration
Not billed
Credit rate1 credit = $0.005

1,000 credits equal $5.00. Token charges are converted from USD pricing into credits and deducted from your account balance.

Every model. One balance.

All prices are in USD per 1 million tokens. Long-context rates apply to the entire request once the prompt exceeds 200K tokens.

Coding · Agents · Documents

Fusion Max 2.0

High-performance coding, agents, and document analysis, built for cost efficiency.

fusion-max-2.0
Context window
1M tokens
Maximum output
128K tokens
Fusion Max 2.0 token rates
Context tierInputCached inputOutput
StandardPROMPT ≤ 200K$3.00/ 1M$0.30/ 1M$20.00/ 1M
Long contextPROMPT > 200K$6.00/ 1M$0.60/ 1M$30.00/ 1M

Reasoning · Agents

Fusion Max 1.0

High-performance reasoning and agents, built for cost efficiency.

fusion-max-1.0
Context window
1M tokens
Maximum output
128K tokens
Fusion Max 1.0 token rates
Context tierInputCached inputOutput
StandardPROMPT ≤ 200K$3.00/ 1M$0.30/ 1M$20.00/ 1M
Long contextPROMPT > 200K$4.50/ 1M$0.45/ 1M$30.00/ 1M

Specialist · Software engineering

Fusion Code 1.0

Cost-efficient coding for software agents.

fusion-code-1.0
Context window
512K tokens
Maximum output
128K tokens
Fusion Code 1.0 token rates
Context tierInputCached inputOutput
StandardPROMPT ≤ 200K$2.00/ 1M$0.50/ 1M$6.00/ 1M
Long contextPROMPT > 200K$4.00/ 1M$1.00/ 1M$12.00/ 1M

Internal collaboration tokens are not billed. Prompt caching is automatic, with no separate cache-write charge.

What counts toward usage.

The API response reports the usage used for billing. Internal model-to-model collaboration remains outside the billable total.

Input tokens
System instructions, user messages, conversation history, tool definitions, tool results, and other context sent with the request.
Cached input
Input served from Prompt Cache is billed at the cached-input rate. Cache misses use the standard input rate.
Output tokens
Final responses, model-generated tool calls, and user-visible reasoning tokens generated by the model.
Fusion collaboration
Tokens generated through FUMO’s internal multi-model collaboration are not billed.

FAQ

How does account balance billing work?+

Add funds before making requests. Each request is priced from its billable input, cached input, and output tokens, converted into credits, then deducted from the balance.

When do long-context rates apply?+

When a prompt exceeds 200K tokens, the long-context rates apply to the entire request, not only the tokens above the threshold.

Is prompt caching automatic?+

Yes. Cache writes incur no separate charge. Cached content may be evicted by the system, so no fixed retention period is guaranteed.

Are the models available through the same API?+

Yes. Fusion models use the same OpenAI-compatible API surface. Select the model ID that fits the work.

Pay as you go

Fund an account. Start with a real workload.

No subscription required. Your balance works across every available Fusion model.

Add balanceContact sales →