Skip to content
InfratoApp
Layers / Inference / Serving / Inference engines / FriendliAI

By FriendliAI

INFERENCE ENGINES

FriendliAI — Client of Friendli Suite. Listed under Inference engines in the Inference / Serving layer. 5 of 9 comparison attributes are recorded, each with the page it was read from.

FriendliAI is the fastest inference cloud for agents, built to run frontier open-weight models in production at scale. It delivers up to 7x faster output token speed, up to 90% lower inference costs, and 99.99% uptime across the most demanding agent workloads — long-context inference, real-time streaming, and accurate tool calling.

2 OF 4 SIGNALS HELD · MISSING LICENCE, REPOSITORY

WebsiteDocsPricing page

What it runs on, and what runs on it

0 cited edges
The differentiator

Computed from the stack graph, both directions. Every cited edge links the page it was read from; a vendor-added one links the vendor’s own source — this is the reason to be here rather than on the vendor’s own site.

No edges on record for FriendliAI.

Nothing in the directory documents depending on it, and its own docs do not name what it runs on. An edge is only stored when a source states it.

Submit a documented integration →

From FriendliAI? No claim is needed for this: a link to your docs that names what FriendliAI runs on is enough.

Open-source signals

LICENCE
not measured
STARS
not measured
DOWNLOADS 30D
457
LAST COMMIT
not measured

A HOLLOW RING IS UNMEASURED, NOT ZERO · WE DO NOT SUBSTITUTE A PARENT COMPANY’S REPOSITORY FOR A PRODUCT THAT HAS NONE

Price

MODELIN /1MCACHED /1MOUT /1MMOVEDTERMSSOURCE
deepseek/deepseek-v3.2$0.50$0.25$1.50— 0.0%$0.50 → $0.50on demand2026-10-21Dispute
google/gemma-4-31b-it$0.14○$0.40— 0.0%$0.14 → $0.14on demand2026-10-21Dispute
minimax/minimax-m2.5$0.30$0.06$1.20— 0.0%$0.30 → $0.30on demand2026-10-21Dispute
z-ai/glm-5.1$1.40$0.26$4.40— 0.0%$1.40 → $1.40on demand2026-10-21Dispute
z-ai/glm-5.2$1.40$0.26$4.40— 0.0%$1.40 → $1.40on demand2026-10-21Dispute
z-ai/glm-5.3$1.40$0.26$4.40— 0.0%$1.40 → $1.40on demand2026-10-21Dispute
z-ai/glm-5.3-flash$0.15$0.03$0.50— 0.0%$0.15 → $0.15on demand2026-10-21Dispute

READINGS OVER 21 DAYS OLD ARE HIDDEN RATHER THAN SHOWN STALE, SO THIS TABLE CAN EMPTY OUT BETWEEN READS · MOVED COMPARES THE CHEAPEST READING WE TOOK ON EACH DAY, WHICH ON A MARKETPLACE IS A DIFFERENT HOST FROM THE ONE PRICED BESIDE IT — BOTH ENDS ARE SHOWN SO NOTHING HAS TO BE INFERRED · A MOVE IS A CHANGE IN OUR READING, WHICH IS NOT ALWAYS A CHANGE IN THE MARKET: A CORRECTED EXTRACTOR MOVES THE SERIES WITHOUT THE PRICE MOVING

Configurations

CONFIGURATIONSWhat you can rent, and what it holds

FriendliAI publishes no machine-readable catalogue of its instance types, so no configurations are recorded. The prices above were read from its pricing page; a configuration read the same way would be a guess about how many GPUs a price covers, which is the mistake this section exists to avoid.

NONE ON RECORD

CONFIGURATIONS ARE TAKEN ONLY FROM A VENDOR’S OWN CATALOGUE API · IF YOU WORK HERE AND PUBLISH ONE, TELL US AND IT WILL BE READ

What its categories ask

5 answered · 4 unanswered

9 questions declared by Inference engines. Most rows across the directory read like these.

ATTRIBUTEANSWEREVIDENCE · SOURCE
Continuous batchingYes
“continuous batching, speculative decoding, and parallel inference — with infrastructure-level optimizations like advanced caching and multi-cloud scaling.”
2026-09-12
Deploymentcloud only2026-09-23
HardwareAscend NPU
“For deployment on the Ascend NPU platform, inference frameworks such as vLLM-Ascend, xLLM and SGLang are supported — see here .”
2026-10-06
Licencemit
“License: mit”
2026-10-06
QuantizationQuantized
“Quantized Deploy”
2026-10-06
Open sourceno answer on record
OpenAI-compatible APIno answer on record
Streamingno answer on record
Structured outputno answer on record

A ROW WITH NO ANSWER IS NOT A ROW WITH A “NO” · A VENDOR MAY ANSWER A BLANK ONE, MARKED VENDOR CLAIMED WITH ITS SOURCE · A DOCUMENTED VALUE IS DISPUTED, NOT OVERWRITTEN, AND STAYS VISIBLE WHILE WE RE-READ THE SOURCE

Practitioner reviews

Nobody has written one yet.

A review here is structured, not a star: hardware, scale band, what broke in production, and whether you would choose it again.

NO PERCENTAGE IS SHOWN UNDER FIVE REVIEWS · NO STAR RATINGS, EVER

Also in this category

47 products in Inference engines. Comparison defaults to within the category — pairing this with a GPU cloud would produce a page that means nothing.

vllmUSED BY 18COMPARE →
atomic-chatNO DEPENDENTS RECORDEDCOMPARE →

USED BY = LISTED PRODUCTS DEPENDING ON IT, IN THIS DIRECTORY

Frequently asked

How much does FriendliAI charge for z-ai/glm-5.3-flash?
$0.03 per token-1m-in-cached, read from FriendliAI's own pricing page on 2026-10-21. 20 rates are listed, each linking to the page it came from.
Does FriendliAI offer continuous batching?
Yes. Read from their own page: “continuous batching, speculative decoding, and parallel inference — with infrastructure-level optimizations like advanced caching and multi-cloud scaling.”.
How current are FriendliAI's prices here?
Every rate on this page was read on 2026-10-21. Nothing older than 21 days is shown: a stale rate is hidden rather than displayed with an old date, and every row links to the page its figure came from.
What does a blank mean on the FriendliAI page?
That nobody has told us, not that the answer is no. A hollow ring is "no answer on record", which is a different state from a recorded no — and the two look different on purpose. A blank is an absence of evidence, and this page shows how many of its questions are in that state rather than quietly leaving them out.
Where do the FriendliAI figures come from?
Each one is read from a public page — usually the vendor's own documentation or pricing page — and carries a link to it with the date it was read. Nothing here is estimated, averaged across sources, or inferred from a similar product. If a figure cannot be traced to a page, it is not published.
Can FriendliAI change what this page says?
They can claim the listing and supply or correct values, which are then labelled as theirs. They cannot remove the page, hide a comparison they are losing, or change a figure somebody else recorded with a source. Nothing on the page is for sale. How the site is paid for is set out under money and independence at /our-data#sec5.
Something here is wrong — how do I fix it?
Dispute the value. Every figure links to the page it was read from, so a dispute names the value and gives a source that contradicts it. That is the only lever anyone has over this page, and it is the same lever for the vendor, a competitor and a reader.
Is this listing run by FriendliAI?
No. It is built from public sources, and no value on it came from the company. Anyone with an address at friendli.ai can claim it for free. A claim lets the company write the tagline, description and logo, and answer blank questions or add dependencies, each marked as theirs with a source. Prices, licence and OSS figures stay as the directory read them, category changes only through our review, and each can be disputed with a source.

Cite this page

A model that quotes us should quote us correctly. The reading and its date sit inside the citation rather than beside it, because a quotation carrying only a URL goes wrong the week the URL changes.

infratoapp, “FriendliAI — Inference engines”, read 21 October 2026. Price $0.50/token-1m-in as published at friendli.ai. https://www.infratoapp.com/products/friendliai

What FriendliAI can and cannot change here

Claiming FriendliAI lets its admins and editors change the tagline, description, logo, docs link and deployment model of every product that org owns — live, labelled as theirs, and logged. Answers to blank questions and dependencies go live too, marked as the owner’s and sourced; a documented answer cannot be overwritten. Everything else on this page is a stated fact with a dispute route: price, licence and OSS figures. Those columns are ungranted at the database level, so they are not a feature we withheld — they are a write that cannot happen. Name, category, website, pricing page and package names can be changed by the owner only through our review. A dispute never deletes a value; it marks it and publishes the correction.

Claim FriendliAIFree · email at friendli.ai