Fish Audio Live 3 与 Grok Voice 3.0 对比：一个价格是套餐，另一个是组件

Put the two rate cards side by side and the newer model looks less than half the price. Fish Audio Live 3 sells for $0.035 per minute on Fish Audio's Agent Platform — a promotional rate, half of a $0.07 list, good through 31 December 2026. G​rok Voice 3.0 sells for $0.08 per minute, or $4.80 an hour, plus $0.004 per text input token. Four cents against eight is a clean story until you notice that the two numbers are billing different objects: G​rok Voice 3.0 is a speech-to-speech component you rent, and Live 3's per-minute figure bundles speech recognition, the reasoning model and the voice into one meter. One is a part, the other is an assembly. Compare them per minute and you are comparing a gearbox to a car.

That distinction is the whole comparison, and it is worth being precise about why. G​rok Voice 3.0 is a versioned model string behind a realtime WebSocket with session-scoped tools and returned transcripts, priced by the minute of audio it processes and scored, independently, at 86.9 on the Artificial Analysis Speech to Speech Index. Fish Audio Live 3 is a full-duplex model launched on 10 November 2026 that keeps listening while it speaks, is English-only at launch by the company's own statement, has no published benchmark of any kind, no model identifier in Fish Audio's API documentation, and no weight release — the vLLM project's public reply to the launch was to hope it gets open-sourced one day.

只存在于一侧的记分牌

Artificial Analysis measures G​rok Voice 3.0 on the same harness it uses for every duplex model, and the numbers are unusually complete. Every figure below is the board's own, taken from its run dated 13 October 2026.

• 语音推理，Big Bench Audio — 98.5%。

• 对话动态，Full Duplex Bench — 97.0%，其中附和反馈 92.9%，打断处理 96.5%，轮次转换 100%，停顿处理 98.5%。

• Speech Agent Arena preference — Elo 1059, against 1074 for G​emini 3.8 Live and 990 for Eleven Realtime v1 Max.

• 任务成功率 — 在 τ-Voice 航空、零售和电信场景中达 94.8%。

• Time to first audio — 0.81 seconds, slower than G​rok Voice 2.0's 0.70 and roughly 40% quicker than GPT-Live-1 in its Astra medium configuration at 1.34 seconds.

• 每小时输入音频的费用——$4.80，直接按供应商自己的每分钟费率换算。

Live 3's column is not sparse, it is empty, and the emptiness has a specific shape. There is no row on that board for any Fish Audio duplex model as of 11 November 2026 — the board lists Fish Audio as a model family and scores none of its live work. Fish Audio's one published number is 186 milliseconds to decide to respond, which the company describes in the context of short reactions such as "mm-hmm" rather than full answers. It is a vendor floor on an acknowledgement. G​rok's 0.81 seconds is an audited time-to-first-audio on a reply. Nothing about those two figures is commensurable, and the fact that both are expressed in seconds is the trap.

The parties' duplex claims actually commit to what

两个模型都是全双工的，因此“边说边听”并不是它们之间的差异化优势；这只是入场券。真正的区别在于时序策略来自哪里。

Fish Audio 的论点是：应答、等待、回答和停止这些行为是从录制的真人对话中学习得来的，而不是写成一份静音阈值清单；它用一个分段读出的电话号码来说明这一点：当人类只听到一次犹豫时，智能体却听到三次话轮结束事件，这正是要修复的失败情形。支撑材料是三段示例通话：完成下单、插话，以及在复述确认过程中进行附和。Fish Audio 尚未公布的是：任何衡量这种学习得到的策略是否优于基于阈值的策略的测量结果。

G​rok Voice 3.0's timing policy is not explained in behavioural terms at all. What xAI published is a rate card, a model string, an alias move — grok-voice-latest now resolves to grok-voice-3.0, and grok-voice-2.0 is flagged deprecated — and, through Artificial Analysis, a set of turn-taking scores that a buyer can read as a risk statement. A 96.5% interruption score is a claim you can test against your own call recordings. A 186-millisecond acknowledgement floor is a claim you have to take on trust until you run it.

The alias move matters more than it sounds in a phone system. G​rok Voice 2.0 became the alias target on 5 August 2026 and was superseded by 3.0 ten weeks later, which means a team that deployed against the alias has already been through two silent model swaps in a quarter. Whether you pin grok-voice-3.0 explicitly is a deployment decision the vendor has effectively forced.

钱到底流向了哪里

捆绑与按组件拆分的划分会产生四种成本形态，其中只有一种在实际通话的检验下仍对 Live 3 有利。

Agent 平台上的 Live 3 —— 促销期间每分钟 $0.035 全包，涵盖识别、推理和语音；电话通话每分钟另加 $0.01，每个电话号码每月 $1.20；若从随附模型升级至高级层 LLM，则按该模型的 token 费率计费。十分钟通话为 35 美分，另加电话费。

• Live 3 after 31 December 2026 — the list rate is $0.07 per minute, at which point the all-in figure is within a cent of G​rok's $0.08 and the bundle argument has to carry the comparison on its own. Fish Audio has not said what happens to the discount.

• G​rok Voice 3.0 — $0.08 per minute of audio plus $0.004 per text input token, with the reasoning model, the telephony and the orchestration all bought separately. The per-minute rate is a floor, not a total.

• 要么，使用你自己的推理模型——这是两者唯一趋同的形态，因为两家供应商都允许你提供大脑，并只对语音层计费。

最后一点，是隐藏在价格表背后的实际结论。如果你的产品是在推理层实现差异化，那么这个套餐只不过是对你本来就要替换掉的东西打个折。如果不是，那么这个套餐确实更便宜、更简单，而 50% 的促销是本月的语音类产品中最划算的交易。

采购，以及报价单无法回答的问题

价格和分数是容易的那一半。困难的那一半在于，当模型在你脚下发生变化时会发生什么，而在这方面，两家厂商做出了截然相反的取舍。

G​rok Voice 3.0 has a stable interface and an unstable model: same endpoint, same event names, same bearer-token handshake, same voice set, with the model swapped behind an alias. Fish Audio Live 3 has an unstable interface and, for now, a stable model: the platform is in public beta, the duplex model is new, its API documentation does not yet name it, and the launch post is the only document that describes its behaviour. One arrangement asks you to test continuously. The other asks you to read release notes.

两者都不算不合理。但在这两者之间做选择的工程团队应当注意，自己更有能力消化哪一种风险，因为相比每分钟四美分，这个答案通常更能预示部署能否成功。

可移植层，以及诚实的边界

Both stacks are two-part systems, and in both the speech half is unportable — G​rok Voice 3.0 runs only on xAI's realtime endpoint, Live 3 only on Fish Audio's platform and on enterprise hardware. The reasoning half is where a router belongs, and on both it is a separate purchase: on the G​rok side it is a model you call from inside the session's tools, on the Fish side it is an OpenAI-compatible endpoint the agent points at. OrcaRouter carries more than 200 models behind one key at provider list price with zero markup, so the text model in either stack can be swapped or pinned without renegotiating the speech contract, and automatic failover keeps a session alive when the reasoning backend errors mid-call. Neither speech model is routable here — there is no third-party endpoint for either — and this piece would be misleading if it implied otherwise.

对于仍在做决定的团队来说，最便宜的实验正是这一定价所促成的那个：用同样的二十通通话分别跑完两者，自己给打断情况打分，再看看别人零售场景上 94.8% 的任务成功率，是否能预测你的场景中的任何东西。Live 3 的 186 毫秒，要么会让你觉得这正是该模型的意义所在，要么就只是一个数字，关乎智能体思考时发出的某种杂音。这不是价目表能回答的问题，而它却是账单到来后唯一重要的问题。