Build Reports disagree 8 publishers2 min readPublished Sep 18, 2026, 2:27 PM PDTUpdated Oct 10, 2026, 7:03 AM PDT

Gemini 3.8 Live runs tool calls in the background without interrupting conversation; Willison's demo UI uses a single WebSocket

Google's two new speech-to-speech models run tool calls in the background without breaking the dialogue, and Artificial Analysis prices them at $0.84 and $3.50 per hour of input audio on one benchmark subset.

The Engineer · Build desk

How we use AI Send a correction

Illustration accompanying Gemini 3.8 Live runs tool calls in the background without interrupting conversation; Willison's demo UI uses a single WebSocket — Generated illustration

What happened

Google released Gemini 3.8 Live and Gemini 3.8 Live Extended Thinking on 15 September 2026, two speech-to-speech models that Simon Willison described as a similar shape to OpenAI's GPT-Live family.
Both models detect and switch between 97 languages mid-conversation, execute tool calls and API requests in the background while they keep talking, and process visual input in near real time.
On Sierra's banking leaderboard for customer-service tasks, Extended Thinking resolves 35.1%, against 32.0% for GPT-Live-1 Astra and 16.5% for xAI-Realtime.
Artificial Analysis's cost per hour of input audio on the Big Bench Audio subset puts 3.8 Live at $0.84 and Extended Thinking at $3.50, against $4.80 for Grok Voice Think Fast 2.0 and $5.83 for GPT-Live-1 Astra.
Simon Willison had OpenAI's GPT-6 Astra Extra High read Google's documentation and build him a browser client for holding voice conversations with the two models.

Why it matters

constraint Budgeting from the $0.84 and $3.50 figures imports Big Bench Audio's ratio of model reasoning to caller speech. A support queue where the agent thinks more per minute of audio pays above that average.
decision The two tiers sit 4.2 times apart in hourly cost and 6.6 points apart in speech quality. Anyone running both has to decide which turns are worth routing to the reasoning tier.
capability A client that is one WebSocket and an AudioContext lets a team instrument its own capture, playback and interrupt timings instead of waiting for an SDK release to expose them.
exposure Every second of audio these models emit carries a SynthID watermark, so recordings a business republishes or archives are detectable as model-generated.

Google's own phrasing for the new capability is real-time reasoning, visual-context handling and background task execution without interrupting the dialogue ^[1]. Neither launch account carries a latency number, so a team rebuilding its voice budget around this has to get the numbers from its own traces.

The transport is documented. Willison's client connects to the BidiGenerateContent WebSocket endpoint and uses a Web Audio API AudioContext for both capture and playback ^[12]. "The implementation uses no libraries," Willison wrote ^[14]. Interrupting the model mid-sentence works in it ^[15]. Capture, playback scheduling and barge-in therefore sit in application code.

Now the two tiers. Extended Thinking scores 82.6% on Artificial Analysis's Speech to Speech Quality Index and the standard model 76.0% ^[5], a gap of 6.6 points ^[18]. Per hour of input audio, Extended Thinking costs 4.2 times the standard tier ^[17]. Measured against GPT-Live-1 Astra's hourly figure, the cheap tier is 6.9 times cheaper ^[25]; officechai.com called it roughly a sixth of the price ^[22].

Against GPT-Live-1 Astra (Medium) the quality lead is 1.1 points on the speech index ^[19] and 0.7 points on Artificial Analysis's agentic benchmark ^[20], while Extended Thinking's hourly cost is 40% lower ^[21].

For an hourly rate to transfer, the traffic has to resemble the subset it was measured on. The denominator is hours of input audio and the numerator covers whatever the model thinks and says in between, so a queue that reasons more per minute of caller speech will not land on the benchmark average. Extended Thinking is the tier sold on multi-step reasoning ^[9]. dev.to reports that the announcement material it saw did not include exact pricing or a full regional rollout schedule ^[26].

The closest thing to a deployment number is Sierra's banking leaderboard, and it leaves 64.9% of tasks unresolved ^[24]. dev.to arrives at the same point in general terms: the practical value will depend on how a business designs its workflows, integrations and escalation paths ^[23].

Availability splits by tier. 3.8 Live goes out through the Gemini API, Google AI Studio and Search Live, with enterprise access through Gemini Enterprise, and Extended Thinking adds Gemini Live plus the Workspace apps Docs, Gmail and Keep for AI Pro and Ultra subscribers ^[6]. dev.to reports the Gemini Enterprise for Customer Experience route as still forthcoming, with enterprise access running through private previews ^[8]. Google names LiveKit, Pipecat and Agora as platform partners and Salesforce and ServiceNow among the enterprise customers building on the models ^[10].

What to watch

A published Google rate card for the Live API, which would replace Artificial Analysis's per-hour estimate with billable per-token terms.
Whether the Gemini Enterprise for Customer Experience route ships, and what escalation hooks it exposes to a support stack.
Independent latency figures: time to first audio, and tool-call turnaround while background reasoning holds the floor.

Clarity's read

What the record supports and how the coverage leans. The claims behind it follow.

Reality

Evidence60
Adoption22
Hype gap+30
Incentives72
Confidence64

Perspective Coverage

8 publishers

Builder: Builder 48%
Operator: Operator 29%
Investor: Investor 23%

Why these scores

Claim ledger

Ranked by verification strength, evidence, and original report placement.

[1]
Google describes the models as capable of real-time reasoning, visual-context handling and background task execution without interrupting the dialogue.
ReportedSupported5 sources— create a free account to open them View cited source
[2]
Both Gemini 3.8 Live models can detect and switch between 97 languages mid-conversation, execute tool calls and API requests in the background while continuing to talk, and process visual input in near real time.
ReportedSupported5 sources— create a free account to open them View cited source
[3]
Google released Gemini 3.8 Live and 3.8 Live Extended Thinking on 15 September 2026, two speech-to-speech models that are a similar shape to OpenAI's GPT-Live family.
ReportedSupported3 sources— create a free account to open them View cited source
[4]
On Artificial Analysis's cost-per-hour-of-input-audio measure using the Big Bench Audio subset, Gemini 3.8 Live costs $0.84 an hour, 3.8 Live Extended Thinking $3.50, Grok Voice Think Fast 2.0 $4.80 and GPT-Live-1 Astra $5.83.
ReportedSupported3 sources— create a free account to open them View cited source
[5]
On Artificial Analysis's Speech to Speech Quality Index, Gemini 3.8 Live Extended Thinking scores 82.6%, GPT-Live-1 Astra (Medium) 81.5%, Grok Voice Think Fast 2.0 (High) 81.3% and the standard Gemini 3.8 Live 76.0%.
ReportedSupported3 sources— create a free account to open them View cited source
[6]
3.8 Live rolls out through the Gemini API, Google AI Studio and Search Live with enterprise access via Gemini Enterprise; Extended Thinking gets the same developer and enterprise rollout plus a consumer release inside Gemini Live and Google Workspace Docs, Gmail and Keep for Google AI Pro and Ultra subscribers.
ReportedSupported4 sources— create a free account to open them View cited source
[7]
On Artificial Analysis's tau-Voice benchmark, 3.8 Live Extended Thinking scores 68.6%, GPT-Live-1 Astra 67.9% and Grok Voice Think Fast 2.0 56.5%.
ReportedSupported3 sources— create a free account to open them View cited source
[8]
dev.to reports that Gemini Enterprise customers get private previews and that the models coming to Gemini Enterprise for Customer Experience is explicitly described as forthcoming.
ReportedSupported4 sources— create a free account to open them View cited source
[9]
Google positions the standard model as a scalable, cost-efficient option for live dialogue and the Extended Thinking version for more complex work that benefits from increased intelligence and multi-step reasoning.
ReportedSupported3 sources— create a free account to open them View cited source
[10]
Google says it is partnering with platforms including LiveKit, Pipecat and Agora, along with enterprise customers including Salesforce and ServiceNow, to build voice agents on the new models.
ReportedSupported3 sources— create a free account to open them View cited source
[11]
Simon Willison pointed GPT-6 Astra Extra High at Google's documentation and had it build a web UI for trying the new models, with model and voice preset selection, an optional system prompt, and a voice conversation in the browser.
ReportedSupported2 sources— create a free account to open them View cited source
[12]
The UI connects to the wss://generativelanguage.googleapis.com BidiGenerateContent WebSocket endpoint and uses a Web Audio API AudioContext for both capture and playback.
ReportedSupported2 sources— create a free account to open them View cited source
[13]
All audio output from the models carries Google's SynthID watermark, its technology for helping detect AI-generated content.
ReportedSupported3 sources— create a free account to open them View cited source
[14]
"The implementation uses no libraries," Willison wrote.
ReportedSupported2 sources— create a free account to open them View cited source
[15]
The browser UI includes the ability to interrupt the model while it is talking.
ReportedSupported2 sources— create a free account to open them View cited source
[16]
On Sierra's tau-cubed Banking Leaderboard, which tests how well a voice agent resolves customer-service style banking tasks, 3.8 Live Extended Thinking leads with 35.1%, against 32.0% for GPT-Live-1 Astra and 16.5% for xAI-Realtime.
ReportedSupported2 sources— create a free account to open them View cited source
[17]
Extended Thinking costs about 4.2 times the standard tier per hour of input audio.
DerivedSupported3 sources— create a free account to open them
[18]
The Speech to Speech Quality Index gap between Extended Thinking and the standard model is 6.6 points.
DerivedSupported3 sources— create a free account to open them
[19]
Extended Thinking's lead over GPT-Live-1 Astra (Medium) on the Speech to Speech Quality Index is 1.1 points.
DerivedSupported3 sources— create a free account to open them
[20]
Extended Thinking's lead over GPT-Live-1 Astra on the tau-Voice benchmark is 0.7 points.
DerivedSupported3 sources— create a free account to open them
[21]
Extended Thinking's hourly input-audio cost is 40% lower than GPT-Live-1 Astra's.
DerivedSupported3 sources— create a free account to open them
[22]
officechai.com wrote that the standard Gemini 3.8 Live model is roughly a sixth of the price of GPT-Live-1 Astra for a good chunk of the same conversational ability.
ReportedSupportedView cited source
[23]
dev.to writes that the practical value will depend on how a business designs its workflows, integrations and escalation paths.
ReportedSupportedView cited source
[24]
Extended Thinking's 35.1% score on Sierra's banking leaderboard leaves 64.9% of tasks unresolved.
DerivedSupported2 sources— create a free account to open them
[25]
Gemini 3.8 Live's hourly cost is 6.9 times lower than GPT-Live-1 Astra's, closer to a seventh than a sixth.
DerivedSupported1 source— create a free account to open them
[26]
dev.to reports that Google has not provided exact pricing or a full regional rollout schedule in the material it saw.
ReportedContested4 sources— create a free account to open them View cited source

Sources

8 independent publishers whose own reporting we read for this story.

blog.google
3 articles · September 15, 2026
Introducing Gemini 3.8 Live and 3.8 Live Extended Thinking
dev.to
2 articles · September 18, 2026
Build real-time voice applications with Gemini 3.8 Live and 3.5 Transcribe
officechai.com
1 article · September 17, 2026
Google Releases Gemini 3.8 Live-Extended Conversational Model, Claims Better Performance Than Rivals At Lower Price
runtimewire.com
1 article · September 15, 2026
Google ships Gemini 3.8 Live to keep voice agents talking while tools work
simonwillison.net
2 articles · September 15, 2026
Gemini Live audio
testingcatalog.com
2 articles · September 15, 2026
Google rolled out Gemini 3.8 Live and Extended Thinking
the-decoder.com
4 articles · September 15, 2026
Apple brings a fully revamped Siri built on Google's Gemini, but not to the EU
thenewstack.io
2 articles · September 15, 2026
OpenAI’s voice model doesn’t think. That’s the point.

Ask Clarity

Have a question this piece left open? The answer comes from its reporting and claim ledger — nothing else.

Try

Clarity analysis

How the coverage leans and what the record actually supports. Read the methodology

Perspective Coverage

How much of the coverage speaks to builders, operators and investors. It measures emphasis, not whether the story is right.

Builder48%

For developers, the useful detail is in Runtime Wire and The New Stack. Extended Thinking rejects synchronous function calls, and a finished spoken sentence can be only a progress update, so an app has to watch a separate interaction_status field to know when work is done. Simon Willison's browser client shows the transport is simple: one WebSocket to the BidiGenerateContent endpoint plus the Web Audio API. "The implementation uses no libraries," he wrote. Runtime Wire adds that the model IDs are listed as stable while the Live API around them still carries preview status.

Reality

Evidence, adoption, hype gap, incentives and confidence, each read independently from the current record on the scale shown beneath it. How these are measured.

Evidence60: 0100
Leaderboard-backed, vendor-relayed
The scores come from named third-party leaderboards at Artificial Analysis and Sierra. Every outlet received them through Google or OfficeChai, and none re-ran a test. Simon Willison's client is the one hands-on check, and it covers conversation and interruption only. Background tool calls, the headline feature, are described in our coverage but not tested.
Adoption22: 0100
Available widely, used by few on record
On paper, distribution is broad on day one, covering developer, consumer and enterprise channels. Enterprise access is still a private preview, though. Evidence of use is limited to Google's partner list and one developer's demo client. Our coverage includes no customer counts or usage figures.
Hype gap+30: −100 understated0 aligned+100 overstated
Quality lead overstated, price lead holds
Google's "#1" framing rests on leads of 1.1 points over GPT-Live-1 Astra on the quality index and 0.7 on tau-Voice. Its banking win still leaves 64.9% of tasks unresolved. The price gap is larger and appears in several outlets' figures. The New Stack points out that Google and OpenAI benchmarked on different tests and setups.
Incentives72: 0100
Google picked the scoreboard
Google selected the leaderboards it cites and wrote the comparison framing. OfficeChai's headline and TestingCatalog's account follow that framing closely. The counterweights are The Decoder's view that OpenAI's model sounds better and The New Stack's caution that the vendors' results are not head-to-head.
Confidence64: 0100
Firm on facts, soft on rankings
Release date, rollout channels, list prices and SynthID match across Google's posts and independent outlets. The comparative claims are weaker. Rival scores and hourly costs come through OfficeChai alone. dev.to's statement that pricing was unpublished conflicts with the per-minute rates in Google's own developer post.

Topics and entities

Follow any of these and your For You feed starts watching them — no settings page required.

Topics

Voice agents and live captioning Follow
Speech-to-speech modelsFollow
Model pricing and inference economics Follow

Entities

OpenAI Follow
Google Follow
Gemini 3.8 Live Extended Thinking Follow
Grok Voice Think Fast 2.0Follow
SynthID Follow
GPT-Live-1Follow
Gemini 3.8 Live Follow
Google DeepMind Follow
Artificial Analysis Follow
Gemini Live APIFollow
Simon Willison Follow
Sierra Follow