Released September 2, 2026, Google's third Flash release in about six weeks (3.6 Flash July 21, 3.7 Flash August 13). Gateway id google/gemini-3.8-flash. $0.75/$3.75 per 1M; Flex $0.375/$1.875. 1,000,000 context, 65,535 max output; text, image, PDF and video input. Google launch figures: Terminal-Bench 2.1 90.8 (3.7 Flash 81.6); HLE-Verified 54.9. A Gemini 3.8 Flash Cyber variant is restricted to governments and vetted partners. Verified October 10, 2026.
Same $0.75/$3.75 price as 3.7 Flash with Google-reported gains on coding and agent benchmarks
Artificial Analysis measured about 302 output tokens per second, among the fastest models it tracks
Accepts text, images, PDF and video natively
Weaknesses
Verbose: Artificial Analysis needed 120M output tokens to run its index against a 71M median, so cost per task runs above the sticker price
65K max output — half of what Claude and OpenAI's current models allow
Real-world use cases
What people actually use Gemini 3.8 Flash for.
Terminal and coding agents — Google reports 90.8% on Terminal-Bench 2.1, up from 81.6% for 3.7 Flash
Video, image and PDF understanding in one 1M-context call
Finance and legal agent workflows, where Google reports gains on Vals Finance Agent V2 and Harvey's legal benchmark
How Gemini 3.8 Flash compares
The nearest models people weigh against it, and what actually separates them.
vs Gemini 3.5 Flash — Against Gemini 3.5 Flash (Google), Gemini 3.8 Flash runs about 57% cheaper per token, gives up 1x on context and answers faster. Take Gemini 3.8 Flash unless you specifically need what Gemini 3.5 Flash does better.
vs Gemini 3.6 Flash — Against Gemini 3.6 Flash (Google), Gemini 3.8 Flash lands within a few percent on price, gives up 1x on context and answers faster. Which one wins depends on whether context depth or latency is your constraint.
vs Muse Spark 1.3 — Against Muse Spark 1.3 (Meta), Gemini 3.8 Flash runs about 18% cheaper per token, gives up 1x on context and answers faster. Take Gemini 3.8 Flash unless you specifically need what Muse Spark 1.3 does better.
Ready to try it?
Start using Gemini 3.8 Flash
Fast, low-cost agentic coding and multimodal work, including video input. Start free — no card required.
Google's I/O 2026 headliner — a Flash-tier model that beats Gemini 3.1 Pro on agentic and coding benchmarks while running roughly 4x faster than comparable frontier models.
Verdict
Excellent fast agentic model, superseded by Gemini 3.6 Flash.
Quality score
89%
Pricing
$1.50/1M in
$9.00/1M out
Speed
Fast
4/5 speed
Context
1.0M tokens
Released at Google I/O, May 19, 2026. Batch API half price; context caching $0.15/1M. 65,536-token output limit.
Google's efficiency-focused successor to 3.5 Flash — higher scores on every benchmark Google tested, ~17% fewer output tokens, and cheaper output pricing.
Verdict
Agent-focused Flash with computer use — since followed by 3.7 and 3.8 Flash.
Quality score
90%
Pricing
$0.75/1M in
$3.75/1M out
Speed
Fast
4/5 speed
Context
1.0M tokens
Released July 21, 2026 alongside 3.5 Flash-Lite and the gated 3.5 Flash Cyber. Knowledge cutoff March 2026. Batch $0.75/$3.75; cached input $0.15/1M.
AgenticComputer useEfficientMultimodal1M context
Best for
Cost-efficient long-horizon agents and computer use
Meta's September 2, 2026 update to Muse Spark — a multimodal reasoning model for long-horizon agent and coding work, with a 1M context and native understanding of video, images and documents.
Verdict
Muse Spark with better tool calling and first-try accuracy, same price.
Quality score
86%
Pricing
$1.25/1M in
$4.25/1M out
Speed
Balanced
3/5 speed
Context
1.0M tokens
Released September 2, 2026. Gateway id meta/muse-spark-1.3. $1.25/$4.25 per 1M. 1,048,576 context and max output; text, image and PDF input. No benchmark table published for this point release. Verified October 10, 2026.
MultimodalAgenticLong contextNew
Best for
Coding agents and multimodal workflows that need few turns and clean output
Pricing moves, ranking shifts, and capability updates.
New ModelSep 2, 2026
Gemini 3.8 Flash released — stronger agentic coding at 3.7 Flash's price
Google released Gemini 3.8 Flash on September 2, 2026, its third Flash model in about six weeks, at the same $0.75/$3.75 per million tokens as 3.7 Flash. Google reports 90.8% on Terminal-Bench 2.1 (3.7 Flash: 81.6%) and 54.9% on HLE-Verified. Artificial Analysis measured about 302 output tokens per second but also found it verbose, using 120M output tokens to run its index against a 71M median — so cost per task runs above the per-token price. It accepts text, images, PDF and video, with a 1M-token context and 65K output. A Gemini 3.8 Flash Cyber variant is limited to governments and vetted partners. Verified October 10, 2026.
Gemini 3.8 Flash costs $0.75 per million input tokens and $3.75 per million output tokens on the API, with cached input at $0.075 per million. A month of 10M input and 2M output tokens runs about $15.00 at list price, before any batch or caching discounts.
What is the context window of Gemini 3.8 Flash?
Gemini 3.8 Flash has a 1M tokens context window, with up to 66k tokens of output per response. That is the total of prompt plus response the model can hold in one request.
What is Gemini 3.8 Flash best for?
Gemini 3.8 Flash is best for fast, low-cost agentic coding and multimodal work, including video input. It is a strong fit when that workflow matters more than the tradeoffs around balanced pricing and very fast speed.
When should I avoid Gemini 3.8 Flash?
You need long single responses (65K output cap) or your workload is output-heavy enough that its verbosity erases the per-token saving.
What is a cheaper alternative to Gemini 3.8 Flash?
Claude Sonnet 5.5 (Anthropic) at $2.00/1M/1M input against Gemini 3.8 Flash's $0.75/1M/1M. Near-Opus 5.5 quality on scoped work at half the price. Compare it first if Gemini 3.8 Flash's pricing is the thing stopping you.
What is a faster alternative to Gemini 3.8 Flash?
Gemini 3.5 Flash — fast against Gemini 3.8 Flash's very fast, with 1.0M tokens of context. Worth the swap when response time is what your users notice rather than the last few points of reasoning depth.
Newsletter
Get notified when Gemini 3.8 Flash pricing changes
We track pricing daily. When this model drops or spikes, you'll know first.
No spam. Useful updates only. Affiliate disclosures always clearly labeled.