オープンソース AI の現状と生産性

V1.0 · Recurring · July 2026

A Letter From Our CTO, Raffi Krikorian

“

In New Zealand's far north, a Māori broadcaster trains speech models for te reo — a language too small for any market — under a license that keeps the data with its people. PwC, one of the largest accounting firms in the world, fine-tuned an open model on the language of finance and runs it today for hundreds of clients, on its own hardware, with no per-token meter running. Researchers in Lausanne built an open medical model with the Red Cross, tuned to its humanitarian guidelines, and are preparing clinical trials at home and in Tanzania. In East Africa, farmers diagnose cassava disease with a model that runs on the phone itself, offline, in fields the cloud has never reached. In Switzerland, a public consortium trained a national model on public supercomputers and released all of it: weights, data, training code. None of them asked permission, and none of them could have rented this. They own it — that is the whole idea.

We have been here before. Mozilla exists because one company tried to own the front door to the web, and an open community rose up to make sure it never could. Twenty-five years later, someone is running the same play. We bet on open the first time. Open won. Together, we can do it again.

Our belief is simple: the path forward is competition and interoperability. We believe in a world of many models, standard ways to plug them together, and the freedom to walk away from any vendor at any time. Open has a record here. It grew the pie and let more people own a slice of it.

Read what follows as a map: where open AI is winning — some numbers surprised even us — and where it is exposed. A case that hides its weak points is an advertisement.”

Open weights closed the capability gap while the price of intelligence collapsed.

0%

Capability gap to the top closed models — at parity on coding, behind on reasoning

0×

Fall in GPT-4-class inference cost in 36 months: $20 → $0.40 per 1M tokens

01The current state of open-source AI

Parity reached. The contest is one layer up.

Open weights are no longer a compromise. They are where the work happens: a majority of production tokens now route through them, and the five highest-volume models on OpenRouter are all open. Closed models still lead at the frontier, on reasoning and multimodality, but the frontier is not what most workloads need. Commodity inputs do not hold pricing power. Value moves up, to the agentic harness.

The capability gap: 8.04% → 0.5% → 3.3%

Open-vs-closed gap on Chatbot Arena over 24 months. By August 2024, the gap had collapsed to 0.5%, and in February 2025 DeepSeek-R1 briefly matched the top US model. By March 2026 it had reopened to 3.3% as closed reasoning models pulled ahead. But 3.3% is an average over a jagged frontier: open is at or near parity on coding, instruction-following and general knowledge, while the gap concentrates in reasoning, long-context retrieval and agentic tasks. The question is no longer whether open models are good enough. It's what you need for your workload. Hover the points.

0%2%4%6%8%

8.04%0.5%R1 matches3.3%Jan 2024Aug 2024Feb 2025Mar 2026

Source: Chatbot Arena, Jan 2024 – Mar 2026.

Inference fell 50× in 36 months

GPT-4-equivalent price per 1M tokens — faster than dotcom-era bandwidth or PC-compute price curves. Log scale.

$0.4$2$10$20

$20$0.40Late 202220232024Dec 2025

Sources: Stanford HAI AI Index 2025 (280× GPT-3.5-class drop over 18 months); Epoch AI (9–900× annual decay); Nov 2025 MIT study (5–10×/yr at the frontier, hardware-adjusted).

Open weights win the tokens

The share of tokens routed on OpenRouter through open-weight models grew from a negligible base to a third by late 2025 to a majority by mid-2026.

0%20%40%60%

~33%majorityNov 2024May 2025Nov 2025Jun 2026

Source: OpenRouter 100T-token study (Nov 2024–Nov 2025) and live leaderboard; intermediate points interpolated. By request count, closed US providers still lead — the open lead is a token-volume lead, concentrated in coding and agentic workloads.

OpenRouter live leaderboard — trailing month, tokens routed

The five highest-volume models are all open weights. Anthropic's closed Claude models are the next US-built entrants.

Open weightsClosed

DeepSeek V4 FlashDeepSeek · open

18.4T

MiMo-V2.5Xiaomi · open

14.9T

Hy3 previewTencent · open

14.8T

MiniMax M3MiniMax · open

14.3T

Owl Alphaundisclosed origin · open

11T

Claude Opus 4.7Anthropic · closed

9.02T

DeepSeek V4 ProDeepSeek · open

8.55T

Claude Sonnet 4.6Anthropic · closed

7.33T

Claude Opus 4.8Anthropic · closed

6.18T

DeepSeek V3.2DeepSeek · open

4.31T

By mid-2026 the top nine models route roughly 18T weekly tokens for Chinese-built models against ~5.5T for US-built ones — more than 3:1 (FT analysis). Where developers route by cost, they route to open weights.

Open ships easy.Open deploys hard.

Data from the Mozilla / SlashData 2026 developer survey. Open models lead in adoption: 79% of developers adding AI functionality use them, against 71% for closed, and the two are largely complementary, with half of developers using both. But production is where teams stall: only 51% of open-model teams reach production versus 63% for closed. The gap is operational tooling and trust, not model capability.

Open models lead in adoption, and mostly coexist with closed

Share of developers adding AI functionality to their applications who currently use each model type, and how the two overlap.

Open models

79%

Closed models

71%

How they combine

29%OS only

50%Both

21%CS only

Source: Mozilla / SlashData 2026 developer survey. Open and closed aren't substitutes for most teams: 50% run both, 29% open only, 21% closed only.

Where open adoption peaks, and where closed still edges it

Open-model adoption by region. Greater China and East Asia lead at 89%; South America and Western Europe are the only two regions where closed adoption exceeds open.

Greater China

89%

East Asia

89%

Western Europe

70%

South America

66%

Same survey, by developer region. In South America and Western Europe, and only there, closed-model adoption runs ahead of open.

Production rate by company size

If the gap were about resources, scale would close it, and it doesn't. Closed climbs 54% → 73% with scale. Open barely moves: 53% → 57%.

Closed modelsOpen models

54%

53%

Small (2–50)

66%

55%

Mid-size (51–1,000)

73%

57%

Enterprise (1,001+)

Enterprises can buy their way through closed deployment. Open deployment waits on tooling nobody has finished. Source: Mozilla / SlashData 2026 developer survey.

Why teams churn: challenges with open models

Δ = churned − still using, in percentage points. The biggest gaps (performance, integration, maintenance) are operational, not capability. Hover the bars.

Still using openChurned away

Performance not good enough

+12pp

Integration into existing systems

+11pp

Ongoing maintenance & updates

+10pp

Insufficient documentation

+8pp

Deployment, hosting, scaling

+8pp

Evaluating / comparing models

+8pp

Fine-tuning & customisation

+4pp

Infrastructure / compute costs

0pp

Security, privacy, compliance

0pp

Lack of specialised support

-2pp

Mozilla survey, n=1,410. “What are the main challenges you face when working with open or open-source AI models?”

The same challenges, everywhere: what blocks open by region

Share of current and churned open-model developers naming each challenge, by region. Warmer cells mean more developers blocked. The top rows are operational in every region: infrastructure cost, security and compliance, maintenance, deployment complexity. South Asia leans hardest on security and support; only North America and Greater China have more than 15% reporting no major challenges.

ChallengeW. Europe & IsraelN. AmericaGreater ChinaSouth AsiaEast Asia ex GCS. AmericaE. Europe & CISOceaniaAll

High infrastructure or compute costs25%26%29%28%28%28%29%18%27%

Security, privacy, or compliance concerns20%27%18%39%29%28%25%22%26%

Ongoing maintenance and updates27%26%18%26%20%31%21%25%24%

Complexity of deployment, hosting, or scaling27%24%19%24%11%30%26%25%23%

Lack of specialised support17%16%21%31%24%23%23%32%22%

Difficulty evaluating or comparing models14%17%14%23%16%26%25%18%18%

Difficulty fine-tuning or customising22%18%18%20%11%22%18%12%18%

Difficulty integrating into existing systems19%21%14%20%7%26%19%20%18%

Insufficient documentation or learning resources18%15%15%17%15%20%24%15%17%

Model performance is not good enough18%15%13%22%16%17%19%8%17%

No major challenges9%21%16%5%14%4%8%12%12%

Weighted sample size28627720619216414798391411

Source: Mozilla / SlashData 2026 developer survey (MZCS1). n=1,410 current or churned open-model developers; the Oceania column (n=39) and Eastern Europe & CIS (n=98) fall below reliable thresholds.

02The open-source AI stack

The open stack scores high on capability,low on operations.

Nine layers and 48 components of the stack scored across 10 criteria (1–5). Click a layer to open its components: each carries its own criterion scores, maturity grade, open-vs-closed parity verdict, and surfaces some of its most-starred open-source projects.

Hover any cell for detail.

StrongViable, but fragmented**Early stage

**Strong (≥4.0)

**3.5–3.9

**3.0–3.4

**2.5–2.9

**Weak (<2.5)

the operational gap = standardization + enterprise readiness

Cells are scores per maturity criterion (1–5), ordered strongest to weakest left to right; layer rows are the means of their components. The two coldest columns, standardization and enterprise readiness, repeat down every layer and every component: that repeating cold edge is the operational gap. Source: Mozilla stack map, June 2026 (48 components, 1,361 projects).

03Who's betting on it

Open source is a business model.

Open-weight AI is a commercial market at multi-hundred-billion-dollar scale, built by funded companies and run in production by global enterprises. Databricks crossed a $5.4B run-rate; Mistral scaled 20× to ~$400M ARR in twelve months; DeepSeek reached ~$220M ARR and recently raised $7.4B at a valuation over $50B. Five revenue models are proven at scale: hosted inference, enterprise platforms, on-prem licensing, fine-tuning services, and harness tooling.

The venture-funded open-source ecosystem: total disclosed funding, USD M

Bars grow as you scroll. Color by region of the company.

North AmericaChina**Europe & rest of world

DeepSeekChina · model

$7.4B

Moonshot AI (Kimi)China · model

$3.9B

Mistral AIFrance · model

$3.0B

Reflection AIUSA · model

$2.1B

Cerebras SystemsUSA · compute

$2.1B

CohereCanada · model

$1.7B

Together AIUSA · inference

$1.3B

BasetenUSA · inference

$585M

Black Forest LabsGermany · model

$450M

Hugging FaceUSA · hub

$400M

ModularUSA · tooling

$380M

Fireworks AIUSA · inference

$327M

AnyscaleUSA · tooling

$281M

LangChainUSA · tooling

$260M

Stability AIUK · model

$230M

Selected companies; Zhipu AI and MiniMax went public (HK IPO 2026) with undisclosed totals. Corporate strategics (Nvidia, Salesforce, AMD, Google, IBM, ASML, Tencent, CATL, Schwarz Group) back the same ecosystem across model, inference, and tooling layers.

Financial maturity of the open ecosystem

Funding, valuation and revenue traction for the companies carrying the open stack. The ecosystem has moved from grants to venture scale to public markets.

CompanyHQLayerDisclosed fundingValuationRevenue signalLeading investorsStage

DatabricksUSAEnterprise platform——$5.4B run-rate—Pre-IPO

DeepSeekChinaFrontier open weights$7.4B$50B+~$220M ARRLiang Wenfeng; Tencent; CATL; China National AI FundPrivate

Mistral AIFranceOpen weights + platform$3.05B~$14B (talks at €20B)~$400M ARR, 20× YoYASML; a16z; Lightspeed; NvidiaPrivate

Moonshot AIChinaOpen weights (Kimi)$3.9B——Meituan/Long-Z; Alibaba; Tencent; HongShanPrivate

Zhipu AIChinaOpen weights (GLM)UndisclosedPublic—Public (HK IPO 2026); prior Alibaba, TencentHK IPO 2026

MiniMaxChinaOpen weightsUndisclosedPublic—Public (HK IPO 2026)HK IPO 2026

CohereCanadaEnterprise / on-prem$1.7B—Command A+ open-sourced May 2026Radical Ventures; Nvidia; AMD; Schwarz GroupPrivate

CerebrasUSACompute$2.1B——Fidelity; Atreides; G42; Tiger GlobalPrivate

Reflection AIUSAOpen weights$2.13B——Nvidia; Disruptive; Sequoia; Lightspeed; DST GlobalPrivate

Together AIUSAInference cloud$1.334B——Aramco Ventures; General Catalyst; Prosperity7; NvidiaPrivate

Hugging FaceUSAHub$400M——Salesforce; Google; Nvidia; IBMPrivate

LangChainUSAHarness tooling$260M—126k+ stars, 60% dev shareIVP; Sequoia; Benchmark; CapitalGPrivate

Five revenue models are proven at scale: hosted inference, enterprise platforms, on-prem licensing, fine-tuning services, and harness tooling. “—” = not publicly disclosed.

The metered model breaks at scale

Closed frontier models are sold by the token — and at production scale the meter becomes the problem.

Microsoft

Cancelling most Claude Code licenses by June 30, 2026 after token billing consumed the annual AI budget in months — while exploring a secured, Azure-hosted DeepSeek V4 for its heaviest Copilot workload.

Uber

Exhausted its entire 2026 AI coding budget in four months, before capping spend at $1,500 per tool per employee, per month.

Stripe

Went the other way: cut inference costs 73% serving open models on vLLM — the same 50M daily API calls on one-third the GPU fleet.

Open-weight self-hosting converts a metered, vendor-controlled operating expense into a fixed cost the enterprise owns.

A fifth of the usage, 4% of the revenue

On OpenRouter (May–Sep 2025), closed models held ~80% of usage and ~96% of revenue. Price drives it: at ~90% parity, closed costs ~6× more per call.

Usage

20%

80%

Open routes roughly a fifth of usage (May–Sep 2025 window; nearer a third full-year, a majority by mid-2026)

Revenue

4%

96%

…while capturing ~4% of model-layer revenue. Usage sits with the commoditizing layer; revenue accrues higher up.

~$24.8B

in unrealized annual savings — the Nagle–Yue study for the Linux Foundation's estimate of the open-vs-closed price asymmetry, at ~6× the cost per call for comparable capability

Where developers route by cost, they route to open weights.

04Why it's happening everywhere

Open isn't a vendor choice.It's a sovereignty choice.

More than 70 national AI strategies are live. The strategic question has shifted from whether to have a national AI policy to which layer of the stack a country can own.

Click a marker or a country below.

The case for open is optionality

Optionality stopped being abstract in June 2026, and it stopped being a procurement question. Three days after Claude Fable 5 went on sale, a single government's export order forced Anthropic to cut access for every foreign national on earth. No other capital was consulted. None could have been. Selective compliance was impossible, so the models went dark for everyone at 5:21 p.m. on a Friday. Anyone who had built on that model inherited a shutdown they had no warning of and no part in. A provider can switch off a model. Nobody can switch off a copy already running on a machine you hold, and that holds whether the machine is a startup's server or a national supercomputer. For a company, weights on disk are a hedge. For a state, they are the difference between a policy and a permission.

The strategic case for open is the ability to leave, and the cloud era proved the cost of its absence:

$90–120k**to move one petabyte out of AWS S3

80%of enterprises now repatriating workloads

$3.2M → <$1M37signals' cloud bill after leaving

2.5×what GEICO's cloud costs ran over plan

Closed model APIs reproduce the same trap: build on a proprietary endpoint and you inherit the vendor's pricing changes with no clean exit. Open weights are exit rights.

The largest source of open weights is China. By design.

Cumulative Hugging Face downloads, March 2026:

Alibaba Qwenopen weights · China

942M

Meta Llamaopen weights · USA

476M

In February 2026 Qwen out-downloaded the next eight organizations combined. On OpenRouter, Chinese open-weight models rose from under 2% of tokens in late 2024 to more than 45% of weekly traffic by April 2026, and about 61% among the ten most-used models. DeepSeek reports 26,000+ enterprise accounts; 58% of new AI startups in 2025 included it in their stack, even as at least eight jurisdictions restricted the hosted service. The resolution is architectural: enterprises ban the hosted app and adopt the weights anyway, self-hosted or via Western endpoints.

This is intentional policy. The State Council's "AI Plus" Initiative (Aug 2025) and the national Five-Year Plan (Mar 2026) codify open-source proliferation as a core directive, and releasing public weights doubles as a macro hedge against semiconductor export controls, offloading global inference onto end users' local hardware. Across the Global South the draw is diversification away from US technology monopolies; elsewhere it is purely financial. Even Microsoft is exploring a secured, Azure-hosted DeepSeek V4 for its heaviest Copilot workload.