OpenAI: GPT-6 Astra

GPT-6 Astra is OpenAI's flagship model for demanding end-to-end work. It is suited for advanced analysis, software engineering, deep research, scientific work, and document creation, with particular …

openai/gpt-6-astra-20260903

GPT-6 Astra is OpenAI's flagship model for demanding end-to-end work. It is suited for advanced analysis, software engineering, deep research, scientific work, and document creation, with particular strengths in long-horizon agentic tasks that involve computer and browser use.

Specifications & capabilities

Input
file, image, text
Output
text
Context window · tokens
1050000
Published model ID
gpt-6-astra
Capabilities
fileInput, multimodal, promptCaching, reasoning, structuredOutputs, toolCalling, visionInput, webSearch
Description source ↗

Evaluations

41–60 / 126

LMArena

1,509.989

source id
lmarena
leaderboard name
LMArena Text Style Control Hard Prompts English
dimension
style
metric
Arena rating
rank
23
sample size
4348
evaluated at
2026-11-09T00:00:00.000Z
source model id
GPT-6 Astra (Max)
confidence lower
1500.2703965714234
confidence upper
1519.7072351692452
Original result ↗

LMArena

1,311.386

source id
lmarena
leaderboard name
LMArena Vision Style Control Diagram
dimension
multimodal
metric
Arena rating
rank
24
sample size
1314
evaluated at
2026-11-09T00:00:00.000Z
source model id
GPT-6 Astra (Max)
confidence lower
1294.2856133426951
confidence upper
1328.48696265155
Original result ↗

LMArena

1,296.997

source id
lmarena
leaderboard name
LMArena Vision Style Control Ocr
dimension
multimodal
metric
Arena rating
rank
24
sample size
3792
evaluated at
2026-11-09T00:00:00.000Z
source model id
GPT-6 Astra (Max)
confidence lower
1286.6017574720584
confidence upper
1307.3914377075373
Original result ↗

OpenRouter Benchmarks

69.667%

source id
openRouterBenchmarks
leaderboard name
tau2-bench Verified Airline · Accuracy · Azure
dimension
agentic
metric
tau2-bench Verified Airline · Accuracy
rank
24
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
Azure
attribution
Source: OpenRouter (openrouter.ai), licensed under CC BY 4.0: reuse and republish with attribution to OpenRouter.
last_run_at
2026-10-01T08:05:41.499Z
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
0.6966665000000001
accuracy_std_dev
0.0033334999999999892
cost_per_task_usd
0.62488186
output_tokens_per_task
3927.45
latency_seconds_per_task
152.08162
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
40
benchmark_evaluation
true
openrouter_permaslug
openai/gpt-6-astra-20260903
source model id
openai/gpt-6-astra-20260903
Original result ↗

OpenRouter Benchmarks

67.556%

source id
openRouterBenchmarks
leaderboard name
tau2-bench Verified Airline · Accuracy · Amazon Bedrock
dimension
agentic
metric
tau2-bench Verified Airline · Accuracy
rank
26
sample size
150
aggregation scope
status
published
harness
url
https://github.com/OpenRouterTeam/benchmark-harness
name
OpenRouter benchmark harness
direction
higherIsBetter
run_count
3
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
Amazon Bedrock
attribution
Source: OpenRouter (openrouter.ai), licensed under CC BY 4.0: reuse and republish with attribution to OpenRouter.
last_run_at
2026-10-30T00:17:00.667Z
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
150
source_file
benchmark-leaderboard.json
source_page
https://openrouter.ai/benchmarks/tau2-bench-airline
measurements
accuracy
0.6755556666666667
accuracy_std_dev
0.006285236475706844
cost_per_task_usd
0.63842362
output_tokens_per_task
3898.48
latency_seconds_per_task
88.20135333333333
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
31
benchmark_evaluation
true
openrouter_permaslug
openai/gpt-6-astra-20260903
source model id
openai/gpt-6-astra-20260903
Original result ↗

LMArena

1,455.552

source id
lmarena
leaderboard name
LMArena Text Style Control Industry Entertainment And Sports And Media
dimension
style
metric
Arena rating
rank
27
sample size
4667
evaluated at
2026-11-09T00:00:00.000Z
source model id
GPT-6 Astra (Max)
confidence lower
1445.7982452422286
confidence upper
1465.306035655867
Original result ↗

LMArena

1,486.951

source id
lmarena
leaderboard name
LMArena Text Style Control English
dimension
style
metric
Arena rating
rank
28
sample size
7113
evaluated at
2026-11-09T00:00:00.000Z
source model id
GPT-6 Astra (Max)
confidence lower
1479.1587765571562
confidence upper
1494.7432531784802
Original result ↗

LMArena

1,461.829

source id
lmarena
leaderboard name
LMArena Text Factuality Spanish
dimension
factuality
metric
Arena rating
rank
29
sample size
446
evaluated at
2026-11-09T00:00:00.000Z
source model id
GPT-6 Astra (Max)
confidence lower
1434.7009926959627
confidence upper
1488.9566810131241
Original result ↗

LMArena

1,327.471

source id
lmarena
leaderboard name
LMArena Vision Style Control Chinese
dimension
multimodal
metric
Arena rating
rank
29
sample size
334
evaluated at
2026-11-09T00:00:00.000Z
source model id
GPT-6 Astra (Max)
confidence lower
1294.1562168364885
confidence upper
1360.7851795244173
Original result ↗

LMArena

1,477.839

source id
lmarena
leaderboard name
LMArena Text Style Control Industry Business And Management And Financial Operations
dimension
style
metric
Arena rating
rank
32
sample size
3154
evaluated at
2026-11-09T00:00:00.000Z
source model id
GPT-6 Astra (Max)
confidence lower
1466.57390692983
confidence upper
1489.1046380381233
Original result ↗

LMArena

1,440.743

source id
lmarena
leaderboard name
LMArena Text Style Control Korean
dimension
style
metric
Arena rating
rank
32
sample size
262
evaluated at
2026-11-09T00:00:00.000Z
source model id
GPT-6 Astra (Max)
confidence lower
1403.0184750601466
confidence upper
1478.4676411922965
Original result ↗

LMArena

1,277.918

source id
lmarena
leaderboard name
LMArena Vision Creative Writing Vision
dimension
multimodal
metric
Arena rating
rank
32
sample size
544
evaluated at
2026-11-09T00:00:00.000Z
source model id
GPT-6 Astra (Max)
confidence lower
1250.9212065506308
confidence upper
1304.9137996458328
Original result ↗

LMArena

1,494.389

source id
lmarena
leaderboard name
LMArena Text Coding
dimension
general
metric
Arena rating
rank
33
sample size
4154
evaluated at
2026-11-09T00:00:00.000Z
source model id
GPT-6 Astra (Max)
confidence lower
1484.6072678132718
confidence upper
1504.1711642990456
Original result ↗

LMArena

1,502.499

source id
lmarena
leaderboard name
LMArena Text Factuality Hard Prompts English
dimension
factuality
metric
Arena rating
rank
34
sample size
4251
evaluated at
2026-11-09T00:00:00.000Z
source model id
GPT-6 Astra (Max)
confidence lower
1494.202427513855
confidence upper
1510.7950487153719
Original result ↗

LMArena

1,500.041

source id
lmarena
leaderboard name
LMArena Text Style Control Hard Prompts
dimension
style
metric
Arena rating
rank
34
sample size
11131
evaluated at
2026-11-09T00:00:00.000Z
source model id
GPT-6 Astra (Max)
confidence lower
1493.4192273820017
confidence upper
1506.6622317822662
Original result ↗

LMArena

1,486.291

source id
lmarena
leaderboard name
LMArena Text Style Control Exclude Ties
dimension
style
metric
Arena rating
rank
34
sample size
13616
evaluated at
2026-11-09T00:00:00.000Z
source model id
GPT-6 Astra (Max)
confidence lower
1479.0190694391167
confidence upper
1493.5628038518344
Original result ↗

LMArena

1,473.482

source id
lmarena
leaderboard name
LMArena Text Style Control Instruction Following
dimension
style
metric
Arena rating
rank
34
sample size
5782
evaluated at
2026-11-09T00:00:00.000Z
source model id
GPT-6 Astra (Max)
confidence lower
1465.0204482965573
confidence upper
1481.942771593338
Original result ↗

LMArena

1,464.949

source id
lmarena
leaderboard name
LMArena Text Style Control Industry Writing And Literature And Language
dimension
style
metric
Arena rating
rank
34
sample size
5171
evaluated at
2026-11-09T00:00:00.000Z
source model id
GPT-6 Astra (Max)
confidence lower
1455.775335435892
confidence upper
1474.1233108675738
Original result ↗

LMArena

1,520.72

source id
lmarena
leaderboard name
LMArena Text Style Control Chinese
dimension
style
metric
Arena rating
rank
35
sample size
1425
evaluated at
2026-11-09T00:00:00.000Z
source model id
GPT-6 Astra (Max)
confidence lower
1504.6520556931605
confidence upper
1536.7883998212696
Original result ↗

LMArena

1,494.963

source id
lmarena
leaderboard name
LMArena Text Factuality Hard Prompts
dimension
factuality
metric
Arena rating
rank
35
sample size
11131
evaluated at
2026-11-09T00:00:00.000Z
source model id
GPT-6 Astra (Max)
confidence lower
1489.2807118267892
confidence upper
1500.6444254481685
Original result ↗

Published prices

1–20 / 133

openai/gpt-6-astra ↗

$2.00

Inference provider
cloudflareWorkersAI
Platform
OpenAI
Unit
cache_read_tokens_per_million
Currency
USD
Billing conditions
label
Long-context cached input (per 1M)
Billing label
Long-context cached input (per 1M)
Pricing via
Cloudflare AI Gateway

openai/gpt-6-astra ↗

$1.00

Inference provider
cloudflareWorkersAI
Platform
OpenAI
Unit
cache_read_tokens_per_million
Currency
USD
Billing conditions
label
Short-context cached input (per 1M)
Billing label
Short-context cached input (per 1M)
Pricing via
Cloudflare AI Gateway

openai/gpt-6-astra ↗

$25.00

Inference provider
cloudflareWorkersAI
Platform
OpenAI
Unit
cache_write_tokens_per_million
Currency
USD
Billing conditions
label
Long-context cache write (per 1M)
Billing label
Long-context cache write (per 1M)
Pricing via
Cloudflare AI Gateway

openai/gpt-6-astra ↗

$12.00

Inference provider
cloudflareWorkersAI
Platform
OpenAI
Unit
cache_write_tokens_per_million
Currency
USD
Billing conditions
label
Short-context cache write (per 1M)
Billing label
Short-context cache write (per 1M)
Pricing via
Cloudflare AI Gateway

openai/gpt-6-astra ↗

$20.00

Inference provider
cloudflareWorkersAI
Platform
OpenAI
Unit
input_tokens_per_million
Currency
USD
Billing conditions
label
Long-context input (per 1M)
Billing label
Long-context input (per 1M)
Pricing via
Cloudflare AI Gateway

openai/gpt-6-astra ↗

$10.00

Inference provider
cloudflareWorkersAI
Platform
OpenAI
Unit
input_tokens_per_million
Currency
USD
Billing conditions
label
Short-context input (per 1M)
Billing label
Short-context input (per 1M)
Pricing via
Cloudflare AI Gateway

openai/gpt-6-astra ↗

$75.00

Inference provider
cloudflareWorkersAI
Platform
OpenAI
Unit
output_tokens_per_million
Currency
USD
Billing conditions
label
Long-context output (per 1M)
Billing label
Long-context output (per 1M)
Pricing via
Cloudflare AI Gateway

openai/gpt-6-astra ↗

$50.00

Inference provider
cloudflareWorkersAI
Platform
OpenAI
Unit
output_tokens_per_million
Currency
USD
Billing conditions
label
Short-context output (per 1M)
Billing label
Short-context output (per 1M)
Pricing via
Cloudflare AI Gateway

openai/gpt-6-astra ↗

$1.00

Inference provider
openRouter
Platform
OpenRouter
Unit
cache_read_tokens_per_million
Currency
USD
Billing conditions

openai/gpt-6-astra ↗

$12.50

Inference provider
openRouter
Platform
OpenRouter
Unit
cache_write_tokens_per_million
Currency
USD
Billing conditions

openai/gpt-6-astra ↗

$10.00

Inference provider
openRouter
Platform
OpenRouter
Unit
input_tokens_per_million
Currency
USD
Billing conditions

openai/gpt-6-astra ↗

$50.00

Inference provider
openRouter
Platform
OpenRouter
Unit
output_tokens_per_million
Currency
USD
Billing conditions

openai/gpt-6-astra ↗

$0.01

Inference provider
openRouter
Platform
OpenRouter
Unit
web_search
Currency
USD
Billing conditions

openai/gpt-6-astra:batch ↗

$0.50

Inference provider
openRouter
Platform
OpenRouter
Unit
cache_read_tokens_per_million
Currency
USD
Billing conditions
mode
batch

openai/gpt-6-astra:batch ↗

$6.25

Inference provider
openRouter
Platform
OpenRouter
Unit
cache_write_tokens_per_million
Currency
USD
Billing conditions
mode
batch

openai/gpt-6-astra:batch ↗

$5.00

Inference provider
openRouter
Platform
OpenRouter
Unit
input_tokens_per_million
Currency
USD
Billing conditions
mode
batch

openai/gpt-6-astra:batch ↗

$25.00

Inference provider
openRouter
Platform
OpenRouter
Unit
output_tokens_per_million
Currency
USD
Billing conditions
mode
batch

~openai/gpt-astra-latest ↗

$1.00

Inference provider
openRouter
Platform
OpenRouter
Unit
cache_read_tokens_per_million
Currency
USD
Billing conditions
alias_target
name
OpenAI: GPT-6 Astra
slug
openai/gpt-6-astra
routing_alias
true

~openai/gpt-astra-latest ↗

$12.50

Inference provider
openRouter
Platform
OpenRouter
Unit
cache_write_tokens_per_million
Currency
USD
Billing conditions
alias_target
name
OpenAI: GPT-6 Astra
slug
openai/gpt-6-astra
routing_alias
true

Provider specifications & access

1–5 / 5

gpt-6-astra ↗

source model id
gpt-6-astra
hosting
gateway
platforms
  • Cheaper Inference
context length
1050000
max output tokens
128000
architecture
input_modalities
  • text
  • image
output_modalities
  • text
capabilities
  • reasoning
  • structuredOutputs
  • toolCalling
  • visionInput
  • webSearch

openai ↗

source model id
GPT-6 Astra
hosting
copilot-subscription
platforms
  • GitHub Copilot
architecture
license
CC-BY-4.0
attribution
GitHub and github/docs contributors
plans
max
true
pro
false
business
true
pro_plus
true
enterprise
true
clients
vs
true
cli
true
xcode
true
dotcom
true
vscode
true
eclipse
true
jetbrains
true
auto selection
app
true
cli
true
chat
true
cloud_agent
true
access scope
Published plan/client flags; account settings, organization policy and additional access conditions apply
availability
supported
release status
GA
billing scope
Copilot subscription usage; not an unconditional inference API rate
billing rates
  • tier
    Default
    unit
    tokens_per_million
    input
    10
    output
    50
    currency
    USD
    original
    tier
    Default
    input
    $10.00
    model
    GPT-6 Astra
    output
    $50.00
    category
    Powerful
    provider
    openai
    threshold
    ≤ 272K
    cache_write
    $12.50
    cached_input
    $1.00
    release_status
    GA
    threshold
    ≤ 272K
    cache_write
    12.5
    cached_input
    1
    footnote_restrictions
    false
    requires_subscription
    true
  • tier
    Long context
    unit
    tokens_per_million
    input
    20
    output
    75
    currency
    USD
    original
    tier
    Long context
    input
    $20.00
    model
    GPT-6 Astra
    output
    $75.00
    category
    Powerful
    provider
    openai
    threshold
    > 272K
    cache_write
    $25.00
    cached_input
    $2.00
    release_status
    GA
    threshold
    > 272K
    cache_write
    25
    cached_input
    2
    footnote_restrictions
    false
    requires_subscription
    true
comparison
name
GPT-6 Astra
excels_at
Long-horizon coding tasks with continuous planning, batched diagnosis and verification, and independent result confirmation
task_area
Long-horizon, autonomous coding and agentic tasks
further_reading
[GPT-6 Astra model card](https://deploymentsafety.openai.com/gpt-6-astra/gpt-6-astra.pdf)

GPT-6 Astra ↗

source model id
openai.gpt-6-astra
hosting
bedrock-managed
platforms
  • Amazon Bedrock
architecture
input_modalities
  • text
  • image
output_modalities
  • text
capabilities

    GPT-6 Astra ↗

    source model id
    openai/gpt-6-astra
    hosting
    third-party
    platforms
    • OpenAI
    context length
    1050000
    max output tokens
    128000
    architecture
    input_modalities
    • text
    • image
    output_modalities
    • text
    capabilities

      GPT-6 Astra ↗

      source model id
      openai/gpt-6-astra
      hosting
      vercel-gateway
      platforms
      • Vercel AI Gateway
      context length
      1050000
      max output tokens
      128000
      architecture
      input_modalities
      • text
      • image
      • pdf
      output_modalities
      • text
      capabilities
      • promptCaching
      • reasoning
      • structuredOutputs
      • toolCalling
      • visionInput
      • webSearch
      regions
      • eu
      • us

      Safety assessments

      1–1 / 1

      Foundation Model Transparency Index 2025 ↗

      35 / 100

      source id
      stanfordFMTI
      canonical model id
      openai/gpt-6-astra-20260903
      taxonomy
      transparency
      hazard category
      transparency
      hazard name
      Organization transparency
      policy area
      OpenAI
      assessment type
      organization-level disclosure
      metric
      disclosure_score
      result label
      35 / 100
      sample size
      100
      updated at
      2026-11-09T18:08:05.227Z

      Source specification observations

      1–12 / 12

      cloudflareWorkersAI ↗

      field
      context_length
      value
      1050000
      normalized value
      1050000
      scope
      source-model
      source id
      cloudflareWorkersAI
      confidence
      1

      openRouter ↗

      field
      context_length
      value
      1050000
      normalized value
      1050000
      scope
      source-model
      source id
      openRouter
      confidence
      1

      openRouter ↗

      field
      context_length
      value
      1050000
      normalized value
      1050000
      scope
      source-model
      source id
      openRouter
      confidence
      1

      openRouter ↗

      field
      context_length
      value
      1050000
      normalized value
      1050000
      scope
      source-model
      source id
      openRouter
      confidence
      1

      vercelAIGatewayCatalog ↗

      field
      context_length
      value
      1050000
      normalized value
      1050000
      scope
      source-model
      source id
      vercelAIGatewayCatalog
      confidence
      1

      cloudflareWorkersAI ↗

      field
      knowledge_cutoff
      reported field
      knowledge_cutoff
      value
      April 30, 2026
      scope
      source-model
      source id
      cloudflareWorkersAI
      confidence
      1

      vercelAIGatewayCatalog ↗

      field
      knowledge_cutoff
      reported field
      knowledge_cutoff
      value
      2026-04-30
      normalized value
      2026-04-30
      precision
      day
      scope
      source-model
      source id
      vercelAIGatewayCatalog
      confidence
      1

      cloudflareWorkersAI ↗

      field
      max_output_tokens
      value
      128000
      normalized value
      128000
      scope
      source-model
      source id
      cloudflareWorkersAI
      confidence
      1

      openRouter ↗

      field
      max_output_tokens
      value
      128000
      normalized value
      128000
      scope
      source-model
      source id
      openRouter
      confidence
      1

      openRouter ↗

      field
      max_output_tokens
      value
      128000
      normalized value
      128000
      scope
      source-model
      source id
      openRouter
      confidence
      1

      openRouter ↗

      field
      max_output_tokens
      value
      128000
      normalized value
      128000
      scope
      source-model
      source id
      openRouter
      confidence
      1

      vercelAIGatewayCatalog ↗

      field
      max_output_tokens
      value
      128000
      normalized value
      128000
      scope
      source-model
      source id
      vercelAIGatewayCatalog
      confidence
      1

      Original sources

      1–15 / 15

      OpenRouter · Model page ↗

      OpenRouter · Model page ↗

      Vercel AI Gateway · Model page ↗

      GitHub · Documentation ↗

      GitHub · Source snapshot ↗

      api.cheaperinference.com · Source snapshot ↗

      Amazon Web Services · Model page ↗

      Amazon Bedrock · Model documentation ↗

      Amazon Web Services · Catalog ↗

      Cloudflare AI · Catalog ↗

      GitHub · Catalog ↗

      LMArena · Catalog ↗

      OpenRouter · Catalog ↗

      OpenRouter · Catalog ↗

      Vercel AI Gateway · Catalog ↗

      Source links

      1–10 / 10

      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