MiniMax: MiniMax M2.5

MiniMax-M2.5 is a SOTA large language model designed for real-world productivity. Trained in a diverse range of complex real-world digital working environments, M2.5 builds upon the coding expertise …

minimax/minimax-m2.5-20260211

MiniMax-M2.5 is a SOTA large language model designed for real-world productivity. Trained in a diverse range of complex real-world digital working environments, M2.5 builds upon the coding expertise of M2.1 to extend into general office work, reaching fluency in generating and operating Word, Excel, and Powerpoint files, context switching between diverse software environments, and working across different agent and human teams. Scoring 80.2% on SWE-Bench Verified, 51.3% on Multi-SWE-Bench, and 76.3% on BrowseComp, M2.5 is also more token efficient than previous generations, having been trained to optimize its actions and output through planning.

Specifications & capabilities

Input
text
Output
text
Context window · tokens
198000
Published model ID
MiniMax-M2.5
Capabilities
promptCaching, reasoning, structuredOutputs, toolCalling
Description source ↗

Evaluations

21–26 / 26

OpenRouter Benchmarks

68.333%

source id
openRouterBenchmarks
leaderboard name
tau2-bench Verified Airline · Accuracy · Venice
dimension
agentic
metric
tau2-bench Verified Airline · Accuracy
rank
30
sample size
100
aggregation scope
status
published
harness
url
https://github.com/OpenRouterTeam/benchmark-harness
name
OpenRouter benchmark harness
direction
higherIsBetter
run_count
2
conditions
window
newest 90 days; successful visible nonzero runs
dataset
benchmark-leaderboard
benchmark
id
tau_bench_verified_airline
name
tau2-bench Verified Airline
record_kind
aggregate
provider_name
Venice
attribution
Source: OpenRouter (openrouter.ai), licensed under CC BY 4.0: reuse and republish with attribution to OpenRouter.
last_run_at
2026-08-29T17:28:21.827Z
license_url
https://creativecommons.org/licenses/by/4.0/
methodology
Rows aggregate the benchmark-runs dataset exactly as the leaderboard pages do, over each benchmark's non-zero-score runs from the newest 90 days: accuracy is the task-weighted mean of per-run accuracy, accuracy_std_dev is the population standard deviation of per-run accuracy (null with a single run), and the per-task averages divide summed cost, generation time, and output tokens by summed tasks. Rows are ordered by accuracy descending, then model name.
metric_kind
tau2-bench Verified Airline · Accuracy
rank_method
competition rank among reported evaluations in identical conditions, including unlinked observations
sample_size
100
source_file
benchmark-leaderboard.json
source_page
https://openrouter.ai/benchmarks/tau2-bench-airline
measurements
accuracy
0.6833334999999999
accuracy_std_dev
0.03666649999999999
cost_per_task_usd
0.0168126
output_tokens_per_task
9021.87
latency_seconds_per_task
195.04515
published_at
2026-11-16T01:46:02.653Z
snapshot_date
2026-11-16
source_revision
7ccaf17664252a3053dcb7df8ad0a46e72954a3b4524ddc1c8a7ef8a41b15568
transformations
TensorTrace preserves source results, converts latency to seconds and derives ranks within matching conditions. Individual runs have no derived rank.
metric_population
38
benchmark_evaluation
true
openrouter_permaslug
minimax/minimax-m2.5-20260211
source model id
minimax/minimax-m2.5-20260211
Original result ↗

OpenRouter Benchmarks

68.047%

source id
openRouterBenchmarks
leaderboard name
GPQA Diamond · Accuracy · Venice
dimension
knowledge
metric
GPQA Diamond · Accuracy
rank
34
sample size
314
aggregation scope
status
published
harness
url
https://github.com/OpenRouterTeam/benchmark-harness
name
OpenRouter benchmark harness
direction
higherIsBetter
run_count
2
conditions
window
newest 90 days; successful visible nonzero runs
dataset
benchmark-leaderboard
benchmark
id
gpqa_diamond
name
GPQA Diamond
record_kind
aggregate
provider_name
Venice
attribution
Source: OpenRouter (openrouter.ai), licensed under CC BY 4.0: reuse and republish with attribution to OpenRouter.
last_run_at
2026-08-29T18:51:07.939Z
license_url
https://creativecommons.org/licenses/by/4.0/
methodology
Rows aggregate the benchmark-runs dataset exactly as the leaderboard pages do, over each benchmark's non-zero-score runs from the newest 90 days: accuracy is the task-weighted mean of per-run accuracy, accuracy_std_dev is the population standard deviation of per-run accuracy (null with a single run), and the per-task averages divide summed cost, generation time, and output tokens by summed tasks. Rows are ordered by accuracy descending, then model name.
metric_kind
GPQA Diamond · Accuracy
rank_method
competition rank among reported evaluations in identical conditions, including unlinked observations
sample_size
314
source_file
benchmark-leaderboard.json
source_page
https://openrouter.ai/benchmarks/gpqa-diamond
measurements
accuracy
0.6804673184713376
accuracy_std_dev
0.00007650000000003487
cost_per_task_usd
0.03821310191082802
output_tokens_per_task
40207.68152866242
latency_seconds_per_task
669.0737388535032
published_at
2026-11-16T01:46:02.653Z
snapshot_date
2026-11-16
source_revision
7ccaf17664252a3053dcb7df8ad0a46e72954a3b4524ddc1c8a7ef8a41b15568
transformations
TensorTrace preserves source results, converts latency to seconds and derives ranks within matching conditions. Individual runs have no derived rank.
metric_population
39
benchmark_evaluation
true
openrouter_permaslug
minimax/minimax-m2.5-20260211
source model id
minimax/minimax-m2.5-20260211
Original result ↗

OpenRouter Benchmarks

62.333%

source id
openRouterBenchmarks
leaderboard name
tau2-bench Verified Airline · Accuracy · Novita
dimension
agentic
metric
tau2-bench Verified Airline · Accuracy
rank
40
sample size
100
aggregation scope
status
published
harness
url
https://github.com/OpenRouterTeam/benchmark-harness
name
OpenRouter benchmark harness
direction
higherIsBetter
run_count
2
conditions
window
newest 90 days; successful visible nonzero runs
dataset
benchmark-leaderboard
benchmark
id
tau_bench_verified_airline
name
tau2-bench Verified Airline
record_kind
aggregate
provider_name
Novita
attribution
Source: OpenRouter (openrouter.ai), licensed under CC BY 4.0: reuse and republish with attribution to OpenRouter.
last_run_at
2026-08-29T16:49:08.574Z
license_url
https://creativecommons.org/licenses/by/4.0/
methodology
Rows aggregate the benchmark-runs dataset exactly as the leaderboard pages do, over each benchmark's non-zero-score runs from the newest 90 days: accuracy is the task-weighted mean of per-run accuracy, accuracy_std_dev is the population standard deviation of per-run accuracy (null with a single run), and the per-task averages divide summed cost, generation time, and output tokens by summed tasks. Rows are ordered by accuracy descending, then model name.
metric_kind
tau2-bench Verified Airline · Accuracy
rank_method
competition rank among reported evaluations in identical conditions, including unlinked observations
sample_size
100
source_file
benchmark-leaderboard.json
source_page
https://openrouter.ai/benchmarks/tau2-bench-airline
measurements
accuracy
0.6233335
accuracy_std_dev
0.01666650000000003
cost_per_task_usd
0.02295258
output_tokens_per_task
9602.23
latency_seconds_per_task
131.17099
published_at
2026-11-16T01:46:02.653Z
snapshot_date
2026-11-16
source_revision
7ccaf17664252a3053dcb7df8ad0a46e72954a3b4524ddc1c8a7ef8a41b15568
transformations
TensorTrace preserves source results, converts latency to seconds and derives ranks within matching conditions. Individual runs have no derived rank.
metric_population
52
benchmark_evaluation
true
openrouter_permaslug
minimax/minimax-m2.5-20260211
source model id
minimax/minimax-m2.5-20260211
Original result ↗

OpenRouter Benchmarks

78.872%

source id
openRouterBenchmarks
leaderboard name
GPQA Diamond · Accuracy · auto-routing
dimension
knowledge
metric
GPQA Diamond · Accuracy
rank
82
sample size
396
aggregation scope
status
published
harness
url
https://github.com/OpenRouterTeam/benchmark-harness
name
OpenRouter benchmark harness
direction
higherIsBetter
run_count
2
conditions
window
newest 90 days; successful visible nonzero runs
dataset
benchmark-leaderboard
benchmark
id
gpqa_diamond
name
GPQA Diamond
record_kind
aggregate
provider_name
auto-routing
attribution
Source: OpenRouter (openrouter.ai), licensed under CC BY 4.0: reuse and republish with attribution to OpenRouter.
last_run_at
2026-08-29T18:05:36.579Z
license_url
https://creativecommons.org/licenses/by/4.0/
methodology
Rows aggregate the benchmark-runs dataset exactly as the leaderboard pages do, over each benchmark's non-zero-score runs from the newest 90 days: accuracy is the task-weighted mean of per-run accuracy, accuracy_std_dev is the population standard deviation of per-run accuracy (null with a single run), and the per-task averages divide summed cost, generation time, and output tokens by summed tasks. Rows are ordered by accuracy descending, then model name.
metric_kind
GPQA Diamond · Accuracy
rank_method
competition rank among reported evaluations in identical conditions, including unlinked observations
sample_size
396
source_file
benchmark-leaderboard.json
source_page
https://openrouter.ai/benchmarks/gpqa-diamond
measurements
accuracy
0.7887204999999999
accuracy_std_dev
0.022727499999999956
cost_per_task_usd
0.02925468181818182
output_tokens_per_task
25932.09595959596
latency_seconds_per_task
425.92292171717173
published_at
2026-11-16T01:46:02.653Z
snapshot_date
2026-11-16
source_revision
7ccaf17664252a3053dcb7df8ad0a46e72954a3b4524ddc1c8a7ef8a41b15568
transformations
TensorTrace preserves source results, converts latency to seconds and derives ranks within matching conditions. Individual runs have no derived rank.
metric_population
132
benchmark_evaluation
true
openrouter_permaslug
minimax/minimax-m2.5-20260211
source model id
minimax/minimax-m2.5-20260211
Original result ↗

OpenRouter Benchmarks

67.667%

source id
openRouterBenchmarks
leaderboard name
tau2-bench Verified Airline · Accuracy · auto-routing
dimension
agentic
metric
tau2-bench Verified Airline · Accuracy
rank
83
sample size
100
aggregation scope
status
published
harness
url
https://github.com/OpenRouterTeam/benchmark-harness
name
OpenRouter benchmark harness
direction
higherIsBetter
run_count
2
conditions
window
newest 90 days; successful visible nonzero runs
dataset
benchmark-leaderboard
benchmark
id
tau_bench_verified_airline
name
tau2-bench Verified Airline
record_kind
aggregate
provider_name
auto-routing
attribution
Source: OpenRouter (openrouter.ai), licensed under CC BY 4.0: reuse and republish with attribution to OpenRouter.
last_run_at
2026-08-29T16:11:34.887Z
license_url
https://creativecommons.org/licenses/by/4.0/
methodology
Rows aggregate the benchmark-runs dataset exactly as the leaderboard pages do, over each benchmark's non-zero-score runs from the newest 90 days: accuracy is the task-weighted mean of per-run accuracy, accuracy_std_dev is the population standard deviation of per-run accuracy (null with a single run), and the per-task averages divide summed cost, generation time, and output tokens by summed tasks. Rows are ordered by accuracy descending, then model name.
metric_kind
tau2-bench Verified Airline · Accuracy
rank_method
competition rank among reported evaluations in identical conditions, including unlinked observations
sample_size
100
source_file
benchmark-leaderboard.json
source_page
https://openrouter.ai/benchmarks/tau2-bench-airline
measurements
accuracy
0.676667
accuracy_std_dev
0.030000000000000027
cost_per_task_usd
0.02188514
output_tokens_per_task
10205.01
latency_seconds_per_task
255.89284
published_at
2026-11-16T01:46:02.653Z
snapshot_date
2026-11-16
source_revision
7ccaf17664252a3053dcb7df8ad0a46e72954a3b4524ddc1c8a7ef8a41b15568
transformations
TensorTrace preserves source results, converts latency to seconds and derives ranks within matching conditions. Individual runs have no derived rank.
metric_population
128
benchmark_evaluation
true
openrouter_permaslug
minimax/minimax-m2.5-20260211
source model id
minimax/minimax-m2.5-20260211
Original result ↗

Published prices

1–13 / 13

minimax/minimax-m2.5 ↗

$0.06

Inference provider
openRouter
Platform
OpenRouter
Unit
cache_read_tokens_per_million
Currency
USD
Billing conditions

minimax/minimax-m2.5 ↗

$0.30

Inference provider
openRouter
Platform
OpenRouter
Unit
input_tokens_per_million
Currency
USD
Billing conditions

minimax/minimax-m2.5 ↗

$1.20

Inference provider
openRouter
Platform
OpenRouter
Unit
output_tokens_per_million
Currency
USD
Billing conditions

minimax/minimax-m2.5 ↗

$0.30

Inference provider
vercelAIGatewayCatalog
Platform
bedrock
Unit
input_tokens_per_million
Currency
USD
Billing conditions
endpoint
bedrock | minimax/minimax-m2.5
model_type
language
billing_scope
via Vercel AI Gateway
Billing label
Input tokens / 1M
Pricing via
Vercel AI Gateway

minimax/minimax-m2.5 ↗

$1.20

Inference provider
vercelAIGatewayCatalog
Platform
bedrock
Unit
output_tokens_per_million
Currency
USD
Billing conditions
endpoint
bedrock | minimax/minimax-m2.5
model_type
language
billing_scope
via Vercel AI Gateway
Billing label
Output tokens / 1M
Pricing via
Vercel AI Gateway

minimax/minimax-m2.5 ↗

$0.03

Inference provider
vercelAIGatewayCatalog
Platform
Vercel AI Gateway
Unit
cache_read_tokens_per_million
Currency
USD
Billing conditions
model_type
language
quote_scope
gateway_catalog
billing_scope
Vercel public catalog; endpoint prices may differ
Billing label
Cached input / 1M
Pricing via
Vercel AI Gateway

minimax/minimax-m2.5 ↗

$0.03

Inference provider
vercelAIGatewayCatalog
Platform
Vercel AI Gateway
Unit
cache_read_tokens_per_million
Currency
USD
Billing conditions
model_type
language
quote_scope
gateway_catalog
service_tier
fast
billing_scope
Vercel public catalog; endpoint prices may differ
Billing label
Cached input / 1M
Pricing via
Vercel AI Gateway

minimax/minimax-m2.5 ↗

$0.38

Inference provider
vercelAIGatewayCatalog
Platform
Vercel AI Gateway
Unit
cache_write_tokens_per_million
Currency
USD
Billing conditions
model_type
language
quote_scope
gateway_catalog
billing_scope
Vercel public catalog; endpoint prices may differ
Billing label
Cache write / 1M
Pricing via
Vercel AI Gateway

minimax/minimax-m2.5 ↗

$0.38

Inference provider
vercelAIGatewayCatalog
Platform
Vercel AI Gateway
Unit
cache_write_tokens_per_million
Currency
USD
Billing conditions
model_type
language
quote_scope
gateway_catalog
service_tier
fast
billing_scope
Vercel public catalog; endpoint prices may differ
Billing label
Cache write / 1M
Pricing via
Vercel AI Gateway

minimax/minimax-m2.5 ↗

$0.30

Inference provider
vercelAIGatewayCatalog
Platform
Vercel AI Gateway
Unit
input_tokens_per_million
Currency
USD
Billing conditions
model_type
language
quote_scope
gateway_catalog
billing_scope
Vercel public catalog; endpoint prices may differ
Billing label
Input tokens / 1M
Pricing via
Vercel AI Gateway

minimax/minimax-m2.5 ↗

$0.60

Inference provider
vercelAIGatewayCatalog
Platform
Vercel AI Gateway
Unit
input_tokens_per_million
Currency
USD
Billing conditions
model_type
language
quote_scope
gateway_catalog
service_tier
fast
billing_scope
Vercel public catalog; endpoint prices may differ
Billing label
Input tokens / 1M
Pricing via
Vercel AI Gateway

minimax/minimax-m2.5 ↗

$1.20

Inference provider
vercelAIGatewayCatalog
Platform
Vercel AI Gateway
Unit
output_tokens_per_million
Currency
USD
Billing conditions
model_type
language
quote_scope
gateway_catalog
billing_scope
Vercel public catalog; endpoint prices may differ
Billing label
Output tokens / 1M
Pricing via
Vercel AI Gateway

minimax/minimax-m2.5 ↗

$2.40

Inference provider
vercelAIGatewayCatalog
Platform
Vercel AI Gateway
Unit
output_tokens_per_million
Currency
USD
Billing conditions
model_type
language
quote_scope
gateway_catalog
service_tier
fast
billing_scope
Vercel public catalog; endpoint prices may differ
Billing label
Output tokens / 1M
Pricing via
Vercel AI Gateway

Provider specifications & access

1–10 / 10

MiniMaxAI/MiniMax-M2.5 ↗

source model id
HuggingFace/minimaxai-minimax-m2.5@1
hosting
foundry-managed-compute
platforms
  • Microsoft Foundry
license
modified-mit
architecture
input_modalities
  • text
output_modalities
  • text
capabilities

    MiniMaxAI/MiniMax-M2.5 ↗

    source model id
    HuggingFace/minimaxai-minimax-m2.5@2
    hosting
    foundry-managed-compute
    platforms
    • Microsoft Foundry
    license
    modified-mit
    architecture
    input_modalities
    • text
    output_modalities
    • text
    capabilities

      MiniMaxAI/MiniMax-M2.5 ↗

      source model id
      HuggingFace/minimaxai-minimax-m2.5@3
      hosting
      foundry-managed-compute
      platforms
      • Microsoft Foundry
      license
      modified-mit
      architecture
      input_modalities
      • text
      output_modalities
      • text
      capabilities

        MiniMaxAI/MiniMax-M2.5 ↗

        source model id
        HuggingFace/minimaxai-minimax-m2.5@4
        hosting
        foundry-managed-compute
        platforms
        • Microsoft Foundry
        license
        modified-mit
        architecture
        input_modalities
        • text
        output_modalities
        • text
        capabilities

          MiniMaxAI/MiniMax-M2.5 ↗

          source model id
          HuggingFace/minimaxai-minimax-m2.5@5
          hosting
          foundry-managed-compute
          platforms
          • Microsoft Foundry
          license
          modified-mit
          architecture
          input_modalities
          • text
          output_modalities
          • text
          capabilities

            MiniMax-M2.5 ↗

            source model id
            MiniMaxAI/MiniMax-M2.5
            hosting
            siliconflow-hosted
            platforms
            • SiliconFlow
            architecture
            input_modalities
            • text
            output_modalities
            • text
            capabilities
            • promptCaching
            • structuredOutputs
            • toolCalling

            MiniMax M2.5 ↗

            source model id
            minimax.minimax-m2.5
            hosting
            bedrock-managed
            platforms
            • Amazon Bedrock
            architecture
            input_modalities
            • text
            output_modalities
            • text
            capabilities

              MiniMax M2.5 ↗

              source model id
              minimax/minimax-m2.5
              hosting
              vercel-gateway
              platforms
              • Vercel AI Gateway
              context length
              204800
              max output tokens
              131000
              architecture
              input_modalities
              • text
              output_modalities
              • text
              capabilities
              • promptCaching
              • reasoning
              • structuredOutputs
              • toolCalling

              MiniMaxAI/MiniMax-M2.5 ↗

              source model id
              publishers/hf-minimaxai/models/minimax-m2.5@001
              hosting
              vertex-self-deployed
              platforms
              • Google Cloud Vertex AI
              architecture
              input_modalities
              • text
              output_modalities
              • text
              capabilities

                minimaxai/minimax-m2.5 ↗

                source model id
                publishers/minimaxai/models/minimax-m2@minimax-m2.5
                hosting
                vertex-self-deployed
                platforms
                • Google Cloud Vertex AI
                architecture
                input_modalities
                • text
                output_modalities
                • text
                capabilities

                  Source specification observations

                  1–5 / 5

                  openRouter ↗

                  field
                  context_length
                  value
                  198000
                  normalized value
                  198000
                  scope
                  source-model
                  source id
                  openRouter
                  confidence
                  1

                  siliconFlow ↗

                  field
                  context_length
                  value
                  197K
                  scope
                  source-model
                  source id
                  siliconFlow
                  confidence
                  1

                  vercelAIGatewayCatalog ↗

                  field
                  context_length
                  value
                  204800
                  normalized value
                  204800
                  scope
                  source-model
                  source id
                  vercelAIGatewayCatalog
                  confidence
                  1

                  openRouter ↗

                  field
                  max_output_tokens
                  value
                  176947
                  normalized value
                  176947
                  scope
                  source-model
                  source id
                  openRouter
                  confidence
                  1

                  vercelAIGatewayCatalog ↗

                  field
                  max_output_tokens
                  value
                  131000
                  normalized value
                  131000
                  scope
                  source-model
                  source id
                  vercelAIGatewayCatalog
                  confidence
                  1

                  Original sources

                  1–20 / 28

                  MiniMax · Model reference ↗

                  SiliconFlow · Model page ↗

                  Hugging Face · MiniMaxAI/MiniMax-M2.5 ↗

                  Google Cloud · Model page ↗

                  Google Cloud · Model page ↗

                  Microsoft Foundry · Model page ↗

                  Microsoft Foundry · Model page ↗

                  Microsoft Foundry · Model page ↗

                  Microsoft Foundry · Model page ↗

                  Microsoft Foundry · Model page ↗

                  OpenRouter · Model page ↗

                  SiliconFlow · Playground ↗

                  Vercel AI Gateway · Model page ↗

                  Microsoft Foundry · Model documentation ↗

                  SiliconFlow · API documentation ↗

                  Amazon Web Services · Model page ↗

                  Amazon Bedrock · Model documentation ↗

                  Amazon Web Services · Catalog ↗

                  Epoch AI · Catalog ↗

                  Google Cloud · Catalog ↗

                  Source links

                  1–12 / 12

                  Hugging Face model card

                  1–1 / 1

                  Explore more models

                  Models

                  Loading
                  EVIDENCE / COMPARISON

                  Benchmarks

                  Explore every benchmark

                  Loading

                  These charts contain model-linked results. The charts above also include unlinked source profiles. Compare scores within each evaluation. Model quality, usage, transparency and system performance measure different things.

                  Install TensorTrace