This model is deprecated. We still benchmark its performance, but less often: every 18 hours instead of every 6.
MiniMax has launched a newer release, MiniMax-M2.7. We suggest considering it instead.
MiniMax-M2.5 API Provider Benchmarking & Analysis
This analysis is intended to support you in choosing the best API provider of MiniMax-M2.5 for your use-case.
Lowest Time per Task
Time per Intelligence Index task (minutes)
Total 6 providers
Lowest Cost per Task
Cost per Intelligence Index task
Total 6 providers
MiniMax-M2.5 is available through 6 API providers, each offering different performance characteristics and pricing. Below is a comparison of the key metrics across providers.
Pricing
Cost per Intelligence Index Task: MiniMax-M2.5
Pricing: Cache Hit, Input, and Output: MiniMax-M2.5
Output Speed vs. Cost per Task: MiniMax-M2.5
Speed
Average of the single-request and multi-turn agentic workloads measured
Output Speed: MiniMax-M2.5
Latency
Average of the single-request and multi-turn agentic workloads measured
Time to First Answer Token: MiniMax-M2.5
Reliability
Share of requests the endpoint failed, across all the single-request and multi-turn agentic workloads measured
Endpoint Error Rate: MiniMax-M2.5
Caching
Cache Hit Rate: MiniMax-M2.5
Cost per Task vs. Cache Hit Rate: MiniMax-M2.5
Key Comparison Metrics & API Features
Frequently Asked Questions
Common questions about MiniMax-M2.5 providers
MiniMax-M2.5 is available through 6 API providers: Novita, MiniMax, GMI FP8, Together AI (FP4), DeepInfra (FP8), and DigitalOcean. Each provider offers different performance characteristics and pricing.
MiniMax-M2.5 is currently available through 6 API providers that we benchmark and track.
The providers with the lowest time to first answer token for MiniMax-M2.5 are DigitalOcean (17.20s), MiniMax (23.88s), and GMI FP8 (30.06s). Lower latency means faster initial response time.
The most affordable providers for MiniMax-M2.5 by blended price are DeepInfra (FP8) ($0.17 per 1M tokens), Novita ($0.20 per 1M tokens), and MiniMax ($0.20 per 1M tokens). Blended price uses a 7:2:1 cache hit/input/output token ratio.
The providers with the lowest input token pricing for MiniMax-M2.5 are DeepInfra (FP8) ($0.15 per 1M input tokens), Novita ($0.30 per 1M input tokens), and MiniMax ($0.30 per 1M input tokens).
The providers with the lowest output token pricing for MiniMax-M2.5 are DeepInfra (FP8) ($1.15 per 1M output tokens), Novita ($1.20 per 1M output tokens), and MiniMax ($1.20 per 1M output tokens).
Prices for MiniMax-M2.5 vary up to 1.3x across providers. The most affordable is DeepInfra (FP8) at $0.17 per 1M tokens, while Together AI (FP4) charges $0.22 per 1M tokens.
All 6 providers of MiniMax-M2.5 support JSON mode for structured output.
All 6 providers of MiniMax-M2.5 support function calling (tool use).
When choosing a provider for MiniMax-M2.5, consider: time per task (how quickly the provider completes an Intelligence Index task), latency (for interactive applications requiring quick first responses), pricing (for cost-sensitive workloads), and API features like JSON mode or function calling.
Yes, provider performance can vary over time due to infrastructure changes, load balancing, and updates. We continuously benchmark all providers and display historical performance trends in the "Over Time" charts.
For information about MiniMax-M2.5's intelligence, capabilities, modalities, and how it compares to other models, see the model overview page. View model overview