MiMo-V2.5-Pro
MiMo-V2.5-Pro on Zumik: live pricing, context, and caching, routable by id or alias through one OpenAI-compatible endpoint.
At a glance.
| Provider | Fireworks AI |
| Family | mimo |
| Released | 2026-05 |
| License | Open weights |
| Context window | 256K tokens |
| Max output | 64K tokens |
| Modalities | text |
| Tool calling | Yes |
| Reasoning mode | Yes |
| Caching | automatic |
| Batch discount | No batch tier |
What reuse looks like here.
Pricing, context, and capabilities for MiMo-V2.5-Pro are live, but it is outside the flagship set Zumik benchmarks in depth, so measured reuse, capture, and warm TTFT are not shown yet. Run a workload estimate or route it by id to start collecting traces.
Same OpenAI client, this model.
from openai import OpenAI
client = OpenAI(base_url="https://api.zumik.ai/v1", api_key="zk_live_...")
r = client.responses.create(
model="mimo-v2p5-pro",
input="Draft a fix for the failing test.",
)
print(r.usage.input_tokens_cached) # confirm reuseMiMo-V2.5-Pro, answered.
How much does MiMo-V2.5-Pro cost?
MiMo-V2.5-Pro is an open-weights model routed through Fireworks AI. It is priced on the host's serverless size tier rather than a single published per-token list price, so it shows "—" here until profiled.
What is MiMo-V2.5-Pro's context window?
MiMo-V2.5-Pro supports a 256K-token context window with up to 64K output tokens.
Does MiMo-V2.5-Pro support prompt caching?
Yes. Fireworks AI uses Automatic prompt caching (serverless and dedicated) caching. In the Zumik corpus, MiMo-V2.5-Pro shows a median cache capture of 81% on agent workloads.
Run MiMo-V2.5-Pro with reuse measured.
Point an OpenAI client at Zumik and see exactly how much of this model's input you are reusing.