Skip to content

LMC Capability Drift Tracker

How LLM capabilities spread through the catalog, quarter by quarter. Each curve shows the percentage of models released in a given release cohort that shipped with a given capability. 346 models tracked across 11 release quarters.

Catalog last refreshed 2026-07-14 (auto-updated hourly from live provider APIs).

Reasoning
100%
+46pp vs Y/Y
was 54% in 2025 Q3
Web Search
57%
+40pp vs Y/Y
was 17% in 2025 Q3
Vision
57%
+29pp vs Y/Y
was 28% in 2025 Q3
Function Calling
100%
+11pp vs Y/Y
was 89% in 2025 Q3
Image Output
0%
0pp vs Y/Y
was 0% in 2025 Q3
JSON Mode
86%
-1pp vs Y/Y
was 87% in 2025 Q3

Capability Adoption Curves

Percentage of models released in each quarter that ship with the given capability.

Reasoning
0% 100%
LMMarketCap.com
Vision
25% 57%
LMMarketCap.com
Function Calling
100% 100%
LMMarketCap.com
JSON Mode
75% 86%
LMMarketCap.com
Web Search
25% 57%
LMMarketCap.com
Image Output
0% 0%
LMMarketCap.com

Cohort Capability Matrix

Share of each release-quarter cohort that ships each capability, expressed as a percentage.

CohortNReasoningVisionFunction CallingJSON ModeWeb SearchImage Output
2024 Q140%25%100%75%25%0%
2024 Q290%33%44%56%0%22%
2024 Q3200%20%45%60%0%20%
2024 Q42313%17%52%30%4%9%
2025 Q12730%37%30%63%33%0%
2025 Q23161%61%84%74%42%0%
2025 Q34654%28%89%87%17%0%
2025 Q45171%67%82%86%37%8%
2026 Q15080%58%88%88%30%2%
2026 Q26594%72%92%88%40%6%
2026 Q314100%57%100%86%57%0%

Capability Debuts

First model in the catalog to ship each capability and how far the capability has spread since.

CapabilityFirst modelLatest adoption
Reasoning
Falcon3 10B Instruct100%
Vision
Claude 3 Haiku57%
Function Calling
GPT-3.5 Turbo100%
JSON Mode
GPT-3.5 Turbo86%
Web Search
Claude 3 Haiku57%
Image Output
DALL-E 30%

Latest models shipping reasoning

The fastest-spreading capability over the last year is reasoning. Here is what recently shipped with it.

ModelProviderReleased
GPT-5.6 Luna ProOpenAI2026-07-09
GPT-5.6 LunaOpenAI2026-07-09
GPT-5.6 Terra ProOpenAI2026-07-09
GPT-5.6 TerraOpenAI2026-07-09
GPT-5.6 Sol ProOpenAI2026-07-09
GPT-5.6 SolOpenAI2026-07-09
Grok 4.5xAI2026-07-08
Grok Latest~x-ai2026-07-08
Aion-3.0-Miniaion-labs2026-07-07
Aion-3.0aion-labs2026-07-07
Hy3Tencent2026-07-06
Hy3 (free)Tencent2026-07-06

How the Capability Drift Tracker works

Every model we index has a release timestamp and a binary capability manifest. We bucket models into release-quarter cohorts using that timestamp, then compute the share of each cohort that shipped with a given capability. The result is an adoption curve per capability per quarter.

We exclude cohorts before 2024 Q1 from the visual curves because those quarters carry fewer than a dozen tracked releases each, so a single model can swing the percentage by tens of points. The cohort table below the charts shows all included quarters and their sample sizes.

Capability flags follow provider-declared manifests reconciled against documented API surfaces. A model that can accept images is counted under vision. A model that exposes a tool-calling endpoint is counted under function calling. A model that supports an explicit reasoning mode (visible or hidden thinking tokens) is counted under reasoning, even if the base mode is plain chat.

Frequently Asked Questions

It is a quarter-by-quarter view of how the LLM catalog has adopted each of six core capabilities: reasoning, vision, function calling, JSON mode, web search, and image output. Each cohort is defined by release quarter, and every model in that cohort contributes to the capability percentages for that quarter. This lets us see which capabilities are becoming table-stakes and which are still rare. Across the latest cohort (2026 Q3, N=14), 100% of new releases shipped with reasoning support and 57% shipped with vision input.

Provider roll-up hides the temporal story. A provider that shipped a reasoning model in 2024 Q4 and a non-reasoning model in 2025 Q2 would average out to 50% if we grouped by provider. Release-cohort grouping lets us see the calendar-quarter at which each capability crossed 50% adoption, which is a much more actionable question for buyers planning a rebuild.

Reasoning is the fastest-spreading capability over the last year. It was at 54% one year ago and now sits at 100% of new releases, a gain of 46 percentage points. Extended chain-of-thought / deliberate reasoning mode with visible or hidden thinking tokens.

Yes. The denominator for each cohort is every tracked model released in that quarter, including open source, free, and paid-only. We do not filter by license or pricing because capability adoption is a platform-wide question, and excluding open source would understate how quickly the open ecosystem is catching up on things like reasoning and vision.

Capability Drift Tracker - How LLM Capabilities Have Evolved | LM Market Cap