What changed in AI
Every material change the connectors observe becomes an event with a source. Dates are when the change occurred (effective date when known); hover a date for the observation time. Older events load with a cursor.
1,212 events until 18 Sept 2026 · ordered by occurrence
17 Sept 2026 27
New model: K2 Horizon 7B (Institute of Foundation Models)
artificial_analysisNew model: K2 Horizon 3.7B (Institute of Foundation Models)
artificial_analysisNew model: K2 Horizon 0.9B (Institute of Foundation Models)
artificial_analysisQwen3 235B A22B Instruct 2507: reasoning changed from true to false
ReasoningYes→Noartificial_analysisQwen3.5-2B: reasoning changed from false to true
ReasoningNo→Yesartificial_analysisDeepSeek V3: max output tokens changed from 16000 to 16384
Max output16K tokens→16.4K tokensopenrouterImagen 4: status changed from deprecated to retired
Statusdeprecated→retiredgoogleGLM 5.3: hf inference providers changed from ["baseten", "deepinfra", "fireworks-ai", "novita", "toget… to ["baseten", "deepinfra", "featherless-ai", "fireworks-ai"…
Hf inference providersbaseten, deepinfra, fireworks-ai, novita, together, zai-org→baseten, deepinfra, featherless-ai, fireworks-ai, novita, together, zai-orghuggingfaceGLM 5.3 Flash: hf inference providers changed from ["baseten", "deepinfra", "fireworks-ai", "novita", "toget… to ["baseten", "deepinfra", "featherless-ai", "fireworks-ai"…
Hf inference providersbaseten, deepinfra, fireworks-ai, novita, together, zai-org→baseten, deepinfra, featherless-ai, fireworks-ai, novita, together, zai-orghuggingfaceHy3: hf inference providers changed from ["deepinfra"] to ["deepinfra", "featherless-ai"]
Hf inference providersdeepinfra→deepinfra, featherless-aihuggingfaceDeepSeek-V4.1-Flash: hf inference providers changed from ["baseten", "deepinfra", "fireworks-ai", "novita"] to ["baseten", "deepinfra", "featherless-ai", "fireworks-ai"…
Hf inference providersbaseten, deepinfra, fireworks-ai, novita→baseten, deepinfra, featherless-ai, fireworks-ai, novitahuggingfaceDeepSeek V4 Pro 0423: hf inference providers changed from ["baseten", "deepinfra", "novita", "together"] to ["baseten", "deepinfra", "featherless-ai", "novita", "tog…
Hf inference providersbaseten, deepinfra, novita, together→baseten, deepinfra, featherless-ai, novita, togetherhuggingfaceDeepSeek V4 Flash (0731): hf inference providers changed from ["baseten", "deepinfra", "fireworks-ai", "novita", "scale… to ["baseten", "deepinfra", "featherless-ai", "fireworks-ai"…
Hf inference providersbaseten, deepinfra, fireworks-ai, novita, scaleway, together→baseten, deepinfra, featherless-ai, fireworks-ai, novita, scaleway, togetherhuggingfaceHy4 preview: release date changed from 2026-08-28 to 2026-08-27
Release date28 Aug 2026→27 Aug 2026huggingfaceGranite 4.2 8B: release date changed from 2026-08-31 to 2026-08-07
Release date31 Aug 2026→7 Aug 2026huggingfaceNew artifact: bartowski/tencent_UI-Mate-27B-GGUF (bartowski)
huggingfaceGLM 4.6: max output tokens changed from 131072 to 16384
Max output131.1K tokens→16.4K tokensopenroutergpt-5.6-terra: reasoning changed from false to true
ReasoningNo→YesopenrouterGLM 5.3 Flash: supported parameters changed from ["frequency_penalty", "include_reasoning", "logit_bias", … to ["frequency_penalty", "include_reasoning", "logit_bias", …
Supported parametersfrequency_penalty, include_reasoning, logit_bias, logprobs, max_tokens, min_p, parallel_tool_calls, presence_penalty, reasoning, reasoning_effort, repetition_penalty, response_format, seed, stop, structured_outputs, temperature, tool_choice, tools, top_k, top_logprobs, top_p→frequency_penalty, include_reasoning, logit_bias, logprobs, max_tokens, min_p, presence_penalty, reasoning, reasoning_effort, repetition_penalty, response_format, seed, stop, structured_outputs, temperature, tool_choice, tools, top_k, top_logprobs, top_popenrouterGLM 4.6: max output tokens changed from 16384 to 131072
Max output16.4K tokens→131.1K tokensopenrouterQwen3.5-4B: reasoning changed from false to true
ReasoningNo→Yesartificial_analysis
16 Sept 2026 21
Cosmos3-Edge: last modified changed from 2026-08-26T21:26:27+00:00 to 2026-09-16T18:48:41+00:00
Last modified2026-08-26T21:26:27+00:00→2026-09-16T18:48:41+00:00huggingfaceClaude Sonnet 4: status changed from deprecated to retired
Statusdeprecated→retiredanthropicClaude Opus 4.1: status changed from deprecated to retired
Statusdeprecated→retiredanthropicLyria 3.5: versions changed from {"preview": ["lyria-3.5"]} to {"stable": ["lyria-3.5"]}
Versions{"preview":["lyria-3.5"]}→{"stable":["lyria-3.5"]}googleQwen 3.8 Max: aa context window changed from 983616 to 1000000
Aa context window983,616→1,000,000artificial_analysisZ.ai GLM 5.2: max output tokens changed from 128000 to 131072
Max output128K tokens→131.1K tokensopenroutercohere-transcribe-03-2026: openness changed from open-source to open-weights
Opennessopen-source→open-weightscohereGLM 5.3 Flash: supported parameters changed from ["frequency_penalty", "include_reasoning", "logit_bias", … to ["frequency_penalty", "include_reasoning", "logit_bias", …
Supported parametersfrequency_penalty, include_reasoning, logit_bias, logprobs, max_tokens, min_p, presence_penalty, reasoning, reasoning_effort, repetition_penalty, response_format, seed, stop, structured_outputs, temperature, tool_choice, tools, top_k, top_logprobs, top_p→frequency_penalty, include_reasoning, logit_bias, logprobs, max_tokens, min_p, parallel_tool_calls, presence_penalty, reasoning, reasoning_effort, repetition_penalty, response_format, seed, stop, structured_outputs, temperature, tool_choice, tools, top_k, top_logprobs, top_popenrouterUnslopNemo 12B: tool calling changed from true to false
Tool callingYes→NoopenrouterUnslopNemo 12B: supported parameters changed from ["frequency_penalty", "logit_bias", "logprobs", "max_toke… to ["frequency_penalty", "logit_bias", "logprobs", "max_toke…
Supported parametersfrequency_penalty, logit_bias, logprobs, max_tokens, presence_penalty, repetition_penalty, response_format, seed, stop, structured_outputs, temperature, tool_choice, tools, top_k, top_logprobs, top_p→frequency_penalty, logit_bias, logprobs, max_tokens, presence_penalty, repetition_penalty, seed, stop, structured_outputs, temperature, top_k, top_logprobs, top_popenrouterQwen3 VL 30B A3B Instruct: max output tokens changed from 32768 to 16384
Max output32.8K tokens→16.4K tokensopenrouterQwen3 VL 30B A3B Instruct: max output tokens changed from 16384 to 32768
Max output16.4K tokens→32.8K tokensopenrouterDeepSeek V3.2: parameter count changed from 685396921376 to 685355329792
Parameters685.4B→685.4BhuggingfaceNemotron 3 Ultra: max output tokens changed from 182520 to 32768
Max output182.5K tokens→32.8K tokensopenrouter
15 Sept 2026 2
gpt-5.5: reasoning changed from false to true
ReasoningNo→YesopenrouterQwen3.8 27B: reasoning changed from false to true
ReasoningNo→Yesopenrouter
Feed keyed on occurred_at. Backfill = history imported when a source is first crawled (excluded — tick the box in the rail to include it). Times are UTC.