Gemini 3.8 Live Extended Thinking
Sep 2026
Gemini 3.8 Live Extended ThinkingCompare inon
Gemini 3.8 Live Extended Thinking measured on 15 of 34 Realtime benchmarks15
BenchmarkByScoreRank
Agents4
Multimodal9
Legacy · saturated or no longer reported2
No benchmark matches
$0.005 / $0.018per minute in / out
1.3 sfirst audio?
68.6τ-Voice
Sep 2026released
en, es, fr +97Languages
Provider-stated input/output audio minute components; text, tools and session/backend charges remain separate.Price covers
Rankingat high effort?
Full ranking →···
15Higgs Realtime92.816Gemini 3.8 Live Extended Thinking91.917GPT Realtime 2.1 Mini91.8Strong at
Alternatives?
ModelScoreCostPriceWhat changes
BetterGPT Live 1low+6%—$0.05Languages: not verified
OpenPersonaPlex 7Bsame——Languages: loses es, fr, de +96
BetterGrok Voice 3.0+5%—$0.08Languages: loses es, zh-Hans, zh-Hant +83
BetterEleven Realtime v1 Mini+4%—$0.08Languages: not verifiedBetterGrok Voice 2.0unstated+3%—$0.08Languages: not verifiedModelScoreCostPriceWhat changes
Nothing here yet.
ModelScoreCostPriceWhat changes
GPT Live 1low+6%—$0.05Languages: not verifiedGrok Voice 3.0+5%—$0.08Languages: loses es, zh-Hans, zh-Hant +83Eleven Realtime v1 Mini+4%—$0.08Languages: not verifiedGrok Voice 2.0unstated+3%—$0.08Languages: not verifiedEleven Realtime v1 Max+2%—$0.08Languages: not verifiedBenchmarks
in Speech & VoiceFull-Duplex-Bench · AA subset v1/v1.591.9#16
Time to first audio · AA1.3#19
AA Speech Agent Arena · overall · Elo979.1#9
AA Speech Agent Arena · overall · Task Success89.1#8
AA Speech Agent Arena · agentic · Elo953.3#11
AA Speech Agent Arena · non-agentic · Elo1,016.6#8
τ-Voice · AA · aggregate68.6#3
τ-Voice · AA · airline72#2
τ-Voice · AA · retail50#4
τ-Voice · AA · telecom83.9#1
EVA-Bench · EVA-A accuracy pass54.8#3
EVA-Bench · EVA-X experience pass73.7#2
EVA-Bench · response speed1.7#3
FAQ
Is Gemini 3.8 Live Extended Thinking free to use?›
There is no free API; the cheapest paid price is $0.005 per 1M input tokens.
How good is Gemini 3.8 Live Extended Thinking, and what is it best at?›
#3 of 17 in Realtime. Strongest at: Speech & Voice.
Is Gemini 3.8 Live Extended Thinking better than Higgs Realtime?›
On Full-Duplex-Bench · AA subset v1/v1.5, Gemini 3.8 Live Extended Thinking scores 91.9 vs 92.8 — about the same.
How can I use Gemini 3.8 Live Extended Thinking?›
Through the Google API.
When was Gemini 3.8 Live Extended Thinking released? Is there a newer version?›
Released 2026-09-15. It is the current version in its line.
Is Gemini 3.8 Live Extended Thinking open source?›
No, Gemini 3.8 Live Extended Thinking is a closed model available through an API. The closest open-weights option here is PersonaPlex 7B.
What is Gemini 3.8 Live Extended Thinking?›
Gemini 3.8 Live Extended Thinking is a closed speech & voice model from Google, released Sep 2026.
What are the best alternatives to Gemini 3.8 Live Extended Thinking?›
GPT Live 1 (better), PersonaPlex 7B (open weights).