Speko Speech-to-Speech Benchmark (six concierge scenarios)
Voice agents book flights, move reservations, and revise budgets during scripted calls. Speko checks whether agents wait for confirmation and use the latest details.
Speko ranks 14 configurations across six scenarios, with five calls per scenario. The overall score combines confirmed actions, task capability, and call completion.
- Made by
- Speko
- Kind
- Customer service
- Level
- Everyday
- Scored
- Checked automatically
- Answers
- Agent task
- Measures
- Capability
- Language
- English
- Interaction
- Agentic
- Environment
- APIs and tools
- Items from
- Written by experts
- Refresh
- One fixed set
- Run by
- A host
- Based on
- A version of another benchmark
- In views
- Agents, Audio and speech
- Last model added
- October 2026
- Scores refreshed
- November 20th, 2026
GPT-Live-1 leads with 0.91 GPT-Live-1 leads with 0.91
Best score per entry from the benchmark's own leaderboard, read on November 20th, 2026. Where a model ran with several settings or harnesses, we keep its best. Names in gray aren't in our model list: fine-tunes, agent systems, products and older models. They don't count toward leaders. This view shows only models we track, each at its best.
- 0.91 GPT-Live-1
- 0.90 Grok Voice 3.0
- 0.87 GPT-Realtime mini
- 0.87 phonic-1-high
- 0.86 GPT-Realtime-2
- 0.84 GPT-Realtime-2.1
- 0.82 gpt-realtime
- 0.82 Grok Voice 2.0
- 0.80 eleven_realtime_v1_max (cascade)
- 0.78 GPT-Realtime-2.1 Mini
- 0.77
Gemini 3.8 Live
- 0.71
Gemini 3.8 Live Extended Thinking
- 0.70 eleven_realtime_v1_mini (cascade)
- 0.67
Gemini 3.1 Flash Live
- 1 GPT-Live-1 0.91
- 2 Grok Voice 3.0 0.90
- 3 GPT-Realtime mini Dec 2025 0.87
- 4 phonic-1-high 0.87
- 5 GPT-Realtime-2 0.86
- 6 GPT-Realtime-2.1 0.84
- 7 gpt-realtime 0.82
- 8 Grok Voice 2.0 0.82
- 9 eleven_realtime_v1_max (cascade) 0.80
- 10 GPT-Realtime-2.1 Mini 0.78
- 11
Gemini 3.8 Live 0.77
- 12
Gemini 3.8 Live Extended Thinking 0.71
- 13 eleven_realtime_v1_mini (cascade) 0.70
- 14
Gemini 3.1 Flash Live 0.67
Source: Official Speko Speech-to-Speech Benchmark (six concierge scenarios) leaderboard
- 1 GPT-Live-1 0.91
- 2 Grok Voice 3.0 0.90
- 3 GPT-Realtime mini Dec 2025 0.87
- 4 GPT-Realtime-2 0.86
- 5 GPT-Realtime-2.1 0.84
- 6 gpt-realtime 0.82
- 7 Grok Voice 2.0 0.82
- 8 GPT-Realtime-2.1 Mini 0.78
- 9
Gemini 3.8 Live 0.77
- 10
Gemini 3.8 Live Extended Thinking 0.71
- 11
Gemini 3.1 Flash Live 0.67
Source: Official Speko Speech-to-Speech Benchmark (six concierge scenarios) leaderboard
| # | Entry | Score ± | Setting | Released | Run on | Source |
|---|---|---|---|---|---|---|
| 1 | GPT-Live-1 | 0.91 0.9 to 0.9 | Jul 2026 | Nov 2026 | Official leaderboard | |
| 2 | Grok Voice 3.0 | 0.90 0.9 to 0.9 | Oct 2026 | Nov 2026 | Official leaderboard | |
| 3 | GPT-Realtime mini Dec 2025 | 0.87 0.8 to 0.9 | Dec 2025 | Nov 2026 | Official leaderboard | |
| 4 | phonic-1-high | 0.87 0.9 to 0.9 | Nov 2026 | Official leaderboard | ||
| 5 | GPT-Realtime-2 | 0.86 ±0 | May 2026 | Nov 2026 | Official leaderboard | |
| 6 | GPT-Realtime-2.1 | 0.84 ±0 | Jul 2026 | Nov 2026 | Official leaderboard | |
| 7 | gpt-realtime | 0.82 ±0 | Aug 2025 | Nov 2026 | Official leaderboard | |
| 8 | Grok Voice 2.0 | 0.82 ±0 | Jul 2026 | Nov 2026 | Official leaderboard | |
| 9 | eleven_realtime_v1_max (cascade) | 0.80 0.8 to 0.9 | Nov 2026 | Official leaderboard | ||
| 10 | GPT-Realtime-2.1 Mini | 0.78 0.7 to 0.8 | Jul 2026 | Nov 2026 | Official leaderboard | |
| 11 | | 0.77 0.7 to 0.8 | Sep 2026 | Nov 2026 | Official leaderboard | |
| 12 | | 0.71 0.7 to 0.8 | Sep 2026 | Nov 2026 | Official leaderboard | |
| 13 | eleven_realtime_v1_mini (cascade) | 0.70 0.7 to 0.7 | Nov 2026 | Official leaderboard | ||
| 14 | | 0.67 ±0 | Mar 2026 | Nov 2026 | Official leaderboard |
| # | Model | Score ± | Setting | Released | Run on | Source |
|---|---|---|---|---|---|---|
| 1 | GPT-Live-1 | 0.91 0.9 to 0.9 | Jul 2026 | Nov 2026 | Official leaderboard | |
| 2 | Grok Voice 3.0 | 0.90 0.9 to 0.9 | Oct 2026 | Nov 2026 | Official leaderboard | |
| 3 | GPT-Realtime mini Dec 2025 | 0.87 0.8 to 0.9 | Dec 2025 | Nov 2026 | Official leaderboard | |
| 4 | GPT-Realtime-2 | 0.86 ±0 | May 2026 | Nov 2026 | Official leaderboard | |
| 5 | GPT-Realtime-2.1 | 0.84 ±0 | Jul 2026 | Nov 2026 | Official leaderboard | |
| 6 | gpt-realtime | 0.82 ±0 | Aug 2025 | Nov 2026 | Official leaderboard | |
| 7 | Grok Voice 2.0 | 0.82 ±0 | Jul 2026 | Nov 2026 | Official leaderboard | |
| 8 | GPT-Realtime-2.1 Mini | 0.78 0.7 to 0.8 | Jul 2026 | Nov 2026 | Official leaderboard | |
| 9 | | 0.77 0.7 to 0.8 | Sep 2026 | Nov 2026 | Official leaderboard | |
| 10 | | 0.71 0.7 to 0.8 | Sep 2026 | Nov 2026 | Official leaderboard | |
| 11 | | 0.67 ±0 | Mar 2026 | Nov 2026 | Official leaderboard |
3 models have held number one on the official leaderboard since 2023
Lab claims count from the day the card came out, independent runs from the model’s release.
Source: Official Speko Speech-to-Speech Benchmark (six concierge scenarios) leaderboard
Scores over time
All the scores we have, one dot each, so a model tested at several settings appears more than once.
Source: Official Speko Speech-to-Speech Benchmark (six concierge scenarios) leaderboard
No lab model card reports it yet
We read the headline results of 540 launch posts from 57 labs, and none of them list this benchmark.
More in Tool use
- AutomationBench In use Agent completes end-to-end business workflows inside a simulated company exposing dozens of REST APIs. Graded on resulting state.
Gemini 4 Neon 56.3%
- MCP Atlas In use Multi-step workflows over real MCP servers. Scores tool-use competency on realistic tasks.
Gemini 4 Argon 90.7%
- BFCL V4 In use Models call functions and tools, including multi-turn agentic tasks with web search and memory. Checked against expected calls and outcomes.
Qwen3.7-Max 75%
- τ³-bench Banking In use Agents act as banking customer-service reps over many turns with a simulated user, following policy. Pass^1 on the final database state.
Qwen 3.8 Max 55.1%
- Toolathlon-Verified In use 108 long-horizon tasks across 32 apps through about 600 MCP tools. Execution-based checkers inspect final environment state. GLM-5.3-Flash 78.4%
- DeepSearchQA In use Questions that need multi-step search and synthesis. Final answers graded against references. Kimi K3 95%
Data audit
8 of 9 checks pass, 1 warning. Look first at the Task access check. We don’t know whether the tasks are public. Every benchmark
Catalog record 1 warning 1 pass
- Warning Task access We don’t know whether the tasks are public.
- Pass Confidence Checked against a primary source.
Scores 3 pass 2 notes
- Pass Independent runs Scores from the official leaderboard (14 entries).
- Pass Models scored 11 models we track, plus 3 other entries on its leaderboard.
- Pass Newest score The newest is from October 2026, for Grok Voice 3.0.
- Note Frontier models None of its models are on the capabilities index.
- Note Headroom Scored as a plain number, with no fixed ceiling.
Leaderboard import 3 pass
- Pass Importer We read the board on speko.com: 11 rows for models we track and 3 other entries at the last read.
- Pass Last read November 20th, 2026.
- Pass Read worked The last read worked.
Consistency 1 pass 1 note
- Note Lab claims No model has both a lab number and an independent run.
- Pass Known issues No published audit found problems.