Google: Gemini 3.8 Flash

Gemini 3.8 Flash is Google's most intelligent Flash model with significant gains from 3.7 Flash across software engineering, agentic tasks, and multi-step reasoning.

google/gemini-3.8-flash-20260902

Gemini 3.8 Flash is Google's most intelligent Flash model with significant gains from 3.7 Flash across software engineering, agentic tasks, and multi-step reasoning.

Specifications & capabilities

Input
text, image, video, file, audio
Output
text
Context window · tokens
1048576
Published model ID
gemini-3.8-flash@default
Capabilities
audioInput, batch, computerUse, fileInput, multimodal, promptCaching, reasoning, structuredOutputs, toolCalling, videoInput, visionInput, webSearch

Evaluations

1–20 / 126

LMArena

1,502.206

source id
lmarena
leaderboard name
LMArena Text Factuality Korean
dimension
factuality
metric
Arena rating
rank
1
sample size
329
evaluated at
2026-11-09T00:00:00.000Z
source model id
Gemini 3.8 Flash (High)
confidence lower
1469.7912733313403
confidence upper
1534.621634482628
Original result ↗

LMArena

1,411.195

source id
lmarena
leaderboard name
LMArena Vision Chinese
dimension
multimodal
metric
Arena rating
rank
1
sample size
364
evaluated at
2026-11-09T00:00:00.000Z
source model id
Gemini 3.8 Flash (High)
confidence lower
1379.3506702367404
confidence upper
1443.0401003992247
Original result ↗

LMArena

1,379.391

source id
lmarena
leaderboard name
LMArena Vision Style Control Chinese
dimension
multimodal
metric
Arena rating
rank
1
sample size
364
evaluated at
2026-11-09T00:00:00.000Z
source model id
Gemini 3.8 Flash (High)
confidence lower
1347.222793385521
confidence upper
1411.5596438432449
Original result ↗

OpenRouter Benchmarks

95.637%

source id
openRouterBenchmarks
leaderboard name
GPQA Diamond · Accuracy · auto-routing
dimension
knowledge
metric
GPQA Diamond · Accuracy
rank
1
sample size
191
aggregation scope
status
published
harness
url
https://github.com/OpenRouterTeam/benchmark-harness
name
OpenRouter benchmark harness
direction
higherIsBetter
run_count
1
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
2026-10-01T08:31:33.481Z
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
191
source_file
benchmark-leaderboard.json
source_page
https://openrouter.ai/benchmarks/gpqa-diamond
measurements
accuracy
0.95637
cost_per_task_usd
0.06822796858638744
output_tokens_per_task
18049.068062827224
latency_seconds_per_task
132.1978062827225
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
google/gemini-3.8-flash-20260902
source model id
google/gemini-3.8-flash-20260902
Original result ↗

OpenRouter Benchmarks

95.286%

source id
openRouterBenchmarks
leaderboard name
GPQA Diamond · Accuracy · Google AI Studio
dimension
knowledge
metric
GPQA Diamond · Accuracy
rank
1
sample size
198
aggregation scope
status
published
harness
url
https://github.com/OpenRouterTeam/benchmark-harness
name
OpenRouter benchmark harness
direction
higherIsBetter
run_count
1
conditions
window
newest 90 days; successful visible nonzero runs
dataset
benchmark-leaderboard
benchmark
id
gpqa_diamond
name
GPQA Diamond
record_kind
aggregate
provider_name
Google AI Studio
attribution
Source: OpenRouter (openrouter.ai), licensed under CC BY 4.0: reuse and republish with attribution to OpenRouter.
last_run_at
2026-10-01T08:29:11.347Z
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
198
source_file
benchmark-leaderboard.json
source_page
https://openrouter.ai/benchmarks/gpqa-diamond
measurements
accuracy
0.952862
cost_per_task_usd
0.07423223737373738
output_tokens_per_task
19646.409090909092
latency_seconds_per_task
86.74828787878788
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
12
benchmark_evaluation
true
openrouter_permaslug
google/gemini-3.8-flash-20260902
source model id
google/gemini-3.8-flash-20260902
Original result ↗

LMArena

1,515.827

source id
lmarena
leaderboard name
LMArena Text Factuality Japanese
dimension
factuality
metric
Arena rating
rank
2
sample size
394
evaluated at
2026-11-09T00:00:00.000Z
source model id
Gemini 3.8 Flash (High)
confidence lower
1489.0774513189972
confidence upper
1542.5772008089448
Original result ↗

LMArena

1,341.6

source id
lmarena
leaderboard name
LMArena Vision Style Control Diagram
dimension
multimodal
metric
Arena rating
rank
2
sample size
1936
evaluated at
2026-11-09T00:00:00.000Z
source model id
Gemini 3.8 Flash (High)
confidence lower
1327.0323428753293
confidence upper
1356.1670163998917
Original result ↗

OpenRouter Benchmarks

73.469%

source id
openRouterBenchmarks
leaderboard name
tau2-bench Verified Airline · Accuracy · Google AI Studio
dimension
agentic
metric
tau2-bench Verified Airline · Accuracy
rank
2
sample size
49
aggregation scope
status
published
harness
url
https://github.com/OpenRouterTeam/benchmark-harness
name
OpenRouter benchmark harness
direction
higherIsBetter
run_count
1
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
Google AI Studio
attribution
Source: OpenRouter (openrouter.ai), licensed under CC BY 4.0: reuse and republish with attribution to OpenRouter.
last_run_at
2026-10-01T08:03:06.861Z
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
49
source_file
benchmark-leaderboard.json
source_page
https://openrouter.ai/benchmarks/tau2-bench-airline
measurements
accuracy
0.734694
cost_per_task_usd
0.25339910204081634
output_tokens_per_task
29831.102040816328
latency_seconds_per_task
146.53116326530613
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
10
benchmark_evaluation
true
openrouter_permaslug
google/gemini-3.8-flash-20260902
source model id
google/gemini-3.8-flash-20260902
Original result ↗

LMArena

1,517.971

source id
lmarena
leaderboard name
LMArena Text Russian
dimension
general
metric
Arena rating
rank
3
sample size
8474
evaluated at
2026-11-09T00:00:00.000Z
source model id
Gemini 3.8 Flash (High)
confidence lower
1510.3087461588727
confidence upper
1525.6331572527215
Original result ↗

LMArena

1,352.924

source id
lmarena
leaderboard name
LMArena Vision Diagram
dimension
multimodal
metric
Arena rating
rank
3
sample size
1936
evaluated at
2026-11-09T00:00:00.000Z
source model id
Gemini 3.8 Flash (High)
confidence lower
1338.4206276575758
confidence upper
1367.4269137299373
Original result ↗

LMArena

1,331.917

source id
lmarena
leaderboard name
LMArena Vision Ocr
dimension
multimodal
metric
Arena rating
rank
3
sample size
5191
evaluated at
2026-11-09T00:00:00.000Z
source model id
Gemini 3.8 Flash (High)
confidence lower
1322.8179992420114
confidence upper
1341.0169791975754
Original result ↗

OpenRouter Benchmarks

94.826%

source id
openRouterBenchmarks
leaderboard name
GPQA Diamond · Accuracy · Google
dimension
knowledge
metric
GPQA Diamond · Accuracy
rank
3
sample size
393
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
Google
attribution
Source: OpenRouter (openrouter.ai), licensed under CC BY 4.0: reuse and republish with attribution to OpenRouter.
last_run_at
2026-10-15T21:00:04.244Z
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
393
source_file
benchmark-leaderboard.json
source_page
https://openrouter.ai/benchmarks/gpqa-diamond
measurements
accuracy
0.9482614885496183
accuracy_std_dev
0.0012429999999999941
cost_per_task_usd
0.07536056743002545
output_tokens_per_task
18975.51145038168
latency_seconds_per_task
132.814941475827
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
google/gemini-3.8-flash-20260902
source model id
google/gemini-3.8-flash-20260902
Original result ↗

LMArena

1,517.364

source id
lmarena
leaderboard name
LMArena Text Style Control Russian
dimension
style
metric
Arena rating
rank
4
sample size
8474
evaluated at
2026-11-09T00:00:00.000Z
source model id
Gemini 3.8 Flash (High)
confidence lower
1509.6911072079408
confidence upper
1525.0377524933483
Original result ↗

LMArena

1,350.904

source id
lmarena
leaderboard name
LMArena Vision Homework
dimension
multimodal
metric
Arena rating
rank
4
sample size
1141
evaluated at
2026-11-09T00:00:00.000Z
source model id
Gemini 3.8 Flash (High)
confidence lower
1332.9497534346247
confidence upper
1368.8588287110763
Original result ↗

LMArena

1,338.363

source id
lmarena
leaderboard name
LMArena Vision Style Control Homework
dimension
multimodal
metric
Arena rating
rank
4
sample size
1141
evaluated at
2026-11-09T00:00:00.000Z
source model id
Gemini 3.8 Flash (High)
confidence lower
1320.360250948317
confidence upper
1356.365405018054
Original result ↗

LMArena

1,328.942

source id
lmarena
leaderboard name
LMArena Vision Humor
dimension
multimodal
metric
Arena rating
rank
4
sample size
299
evaluated at
2026-11-09T00:00:00.000Z
source model id
Gemini 3.8 Flash (High)
confidence lower
1293.627948229118
confidence upper
1364.2555345325422
Original result ↗

LMArena

1,315.428

source id
lmarena
leaderboard name
LMArena Vision Style Control Ocr
dimension
multimodal
metric
Arena rating
rank
4
sample size
5191
evaluated at
2026-11-09T00:00:00.000Z
source model id
Gemini 3.8 Flash (High)
confidence lower
1306.2934860232658
confidence upper
1324.5620476091756
Original result ↗

LMArena

1,304.796

source id
lmarena
leaderboard name
LMArena Vision Style Control Humor
dimension
multimodal
metric
Arena rating
rank
4
sample size
299
evaluated at
2026-11-09T00:00:00.000Z
source model id
Gemini 3.8 Flash (High)
confidence lower
1269.2249876047101
confidence upper
1340.367377204374
Original result ↗

LMArena

1,525.556

source id
lmarena
leaderboard name
LMArena Text Math
dimension
general
metric
Arena rating
rank
5
sample size
3076
evaluated at
2026-11-09T00:00:00.000Z
source model id
Gemini 3.8 Flash (High)
confidence lower
1513.9626369756056
confidence upper
1537.149673791695
Original result ↗

LMArena

1,519.529

source id
lmarena
leaderboard name
LMArena Text Style Control Math
dimension
style
metric
Arena rating
rank
5
sample size
3076
evaluated at
2026-11-09T00:00:00.000Z
source model id
Gemini 3.8 Flash (High)
confidence lower
1507.9447390049022
confidence upper
1531.1139237412967
Original result ↗

Published prices

61–61 / 61

google/gemini-3.8-flash ↗

$0.01

Inference provider
vercelAIGatewayCatalog
Platform
vertex
Unit
request
Currency
USD
Billing conditions
charge
web_search
endpoint
vertex | google/gemini-3.8-flash
model_type
language
billing_scope
via Vercel AI Gateway
Billing label
Web search
Pricing via
Vercel AI Gateway

Provider specifications & access

1–5 / 5

gemini-3.8-flash ↗

source model id
gemini-3.8-flash
hosting
gateway
platforms
  • Cheaper Inference
context length
1048576
max output tokens
65535
architecture
input_modalities
  • text
  • image
  • video
output_modalities
  • text
capabilities
  • reasoning
  • structuredOutputs
  • toolCalling
  • videoInput
  • visionInput
  • webSearch

google ↗

source model id
Gemini 3.8 Flash
hosting
copilot-subscription
platforms
  • GitHub Copilot
architecture
license
CC-BY-4.0
attribution
GitHub and github/docs contributors
plans
max
true
pro
true
business
true
pro_plus
true
enterprise
true
clients
vs
true
cli
true
xcode
true
dotcom
false
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
    0.75
    output
    3.75
    currency
    USD
    original
    tier
    Default
    input
    $0.75
    model
    Gemini 3.8 Flash[^gemini-flash-promo]
    output
    $3.75
    category
    Versatile
    provider
    google
    threshold
    Not applicable
    cached_input
    $0.075
    release_status
    GA
    threshold
    Not applicable
    cached_input
    0.075
    footnote_restrictions
    true
    requires_subscription
    true
comparison
name
Gemini 3.8 Flash
excels_at
Fast, reliable answers to lightweight coding questions
task_area
Fast help with simple or repetitive tasks
further_reading
[Gemini 3.8 Flash model card](https://storage.googleapis.com/deepmind-media/Model-Cards/Gemini-3-8-Flash-Model-Card.pdf)

Gemini 3.8 Flash ↗

source model id
google/gemini-3.8-flash
hosting
third-party
platforms
  • Google Vertex AI
context length
1048576
max output tokens
65536
architecture
input_modalities
  • text
output_modalities
  • text
capabilities

    Gemini 3.8 Flash ↗

    source model id
    google/gemini-3.8-flash
    hosting
    vercel-gateway
    platforms
    • Vercel AI Gateway
    context length
    1000000
    max output tokens
    65535
    architecture
    input_modalities
    • text
    • image
    • pdf
    • video
    output_modalities
    • text
    capabilities
    • promptCaching
    • reasoning
    • structuredOutputs
    • toolCalling
    • videoInput
    • visionInput
    • webSearch
    regions
    • eu
    • us

    Gemini 3.8 Flash ↗

    source model id
    publishers/google/models/gemini-3.8-flash@default
    hosting
    vertex-managed
    platforms
    • Google Cloud Vertex AI
    context length
    1048576
    max output tokens
    65536
    architecture
    input_modalities
    • audio
    • image
    • text
    • video
    output_modalities
    • text
    capabilities
    • batch
    • computerUse
    • multimodal
    • promptCaching
    • reasoning
    • structuredOutputs
    • toolCalling
    • visionInput

    Safety assessments

    1–1 / 1

    Foundation Model Transparency Index 2025 ↗

    41 / 100

    source id
    stanfordFMTI
    canonical model id
    google/gemini-3.8-flash-20260902
    taxonomy
    transparency
    hazard category
    transparency
    hazard name
    Organization transparency
    policy area
    Google
    assessment type
    organization-level disclosure
    metric
    disclosure_score
    result label
    41 / 100
    sample size
    100
    updated at
    2026-11-09T18:08:05.227Z

    Source specification observations

    1–10 / 10

    cloudflareWorkersAI ↗

    field
    context_length
    value
    1048576
    normalized value
    1048576
    scope
    source-model
    source id
    cloudflareWorkersAI
    confidence
    1

    openRouter ↗

    field
    context_length
    value
    1048576
    normalized value
    1048576
    scope
    source-model
    source id
    openRouter
    confidence
    1

    openRouter ↗

    field
    context_length
    value
    1048576
    normalized value
    1048576
    scope
    source-model
    source id
    openRouter
    confidence
    1

    vercelAIGatewayCatalog ↗

    field
    context_length
    value
    1000000
    normalized value
    1000000
    scope
    source-model
    source id
    vercelAIGatewayCatalog
    confidence
    1

    vertexAI ↗

    field
    context_length
    value
    1048576
    normalized value
    1048576
    scope
    source-model
    source id
    vertexAI
    confidence
    1

    cloudflareWorkersAI ↗

    field
    max_output_tokens
    value
    65536
    normalized value
    65536
    scope
    source-model
    source id
    cloudflareWorkersAI
    confidence
    1

    openRouter ↗

    field
    max_output_tokens
    value
    65536
    normalized value
    65536
    scope
    source-model
    source id
    openRouter
    confidence
    1

    openRouter ↗

    field
    max_output_tokens
    value
    65536
    normalized value
    65536
    scope
    source-model
    source id
    openRouter
    confidence
    1

    vercelAIGatewayCatalog ↗

    field
    max_output_tokens
    value
    65535
    normalized value
    65535
    scope
    source-model
    source id
    vercelAIGatewayCatalog
    confidence
    1

    vertexAI ↗

    field
    max_output_tokens
    value
    65536
    normalized value
    65536
    scope
    source-model
    source id
    vertexAI
    confidence
    1

    Original sources

    1–18 / 18

    Google AI · Model reference ↗

    Google Cloud · Model page ↗

    OpenRouter · Model page ↗

    OpenRouter · Model page ↗

    Vercel AI Gateway · Model page ↗

    GitHub · Documentation ↗

    Google Cloud · Documentation ↗

    GitHub · Source snapshot ↗

    api.cheaperinference.com · Source snapshot ↗

    Cloudflare AI · Catalog ↗

    Epoch AI · Catalog ↗

    GitHub · Catalog ↗

    Google Cloud · Catalog ↗

    LMArena · Catalog ↗

    OpenRouter · Catalog ↗

    OpenRouter · Catalog ↗

    Vercel AI Gateway · Catalog ↗

    Vercel AI Gateway · Catalog ↗

    Source links

    1–12 / 12

    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