Benchmark Almanac

Indic-KCC-Agri-Advisory-Benchmark

DeepSeek-V4-Flash 4 leads since September 2026
Not in model cards

Models answer crop questions from India’s Kisan Call Center in English and ten Indian languages. A judge compares each answer with a human advisor’s response.

The maker ranks 21 models on 500 questions in each language. A Qwen judge rates correctness, naturalness, grounding and safety separately from one to five.

Made by
Sthānika AI
Released
September 2026
Kind
Knowledge in many languages
Level
University
Scored
Graded by a model
Answers
Open-ended
Tasks
All public
Measures
Capability
Language
Many languages
In and out
Text → Text
Interaction
Single turn
Items from
Real work and use
Refresh
One fixed set
Run by
Anyone
In views
Many languages
Strongest model tested
DeepSeek-V4-Flash #87 on the capability index
Last model added
April 2026
Scores refreshed
November 20th, 2026
21
Entries on its own leaderboard, 9 models we track

DeepSeek-V4-Flash leads with 4 among models released by April 2026 DeepSeek-V4-Flash leads with 4 among models released by April 2026

Best score per entry from the benchmark's own leaderboard, read on November 20th, 2026. Where a model ran with several settings or harnesses, we keep its best. Names in gray aren't in our model list: fine-tunes, agent systems, products and older models. They don't count toward leaders. This view shows only models we track, each at its best.

The newest model with a score here came out in April 2026, and nobody has tested a newer one. Among the 20 highest-ranked models on the capability index, these have no score yet:

Bars start at 1.

  1. 1 DeepSeek-V4-Flash 4
  2. 2 Gemma 3 27B 3
  3. 3 Qwen3.6 27B 3
  4. 4 mistral-small-3.1-24b (anomaly flagged) 3
  5. 5 gemma3-12b-int4 (QAT) 2
  6. 6 Gemma 3 12B 2
  7. 7 Phi-4 2
  8. 8 gpt-oss-20b 2
  9. 9 llama3.1-8b-instruct 2
  10. 10 Qwen3-VL-8B-Instruct 2
  11. 11 Gemma 3 4B 2
  12. 12 krutrim-1-7b 2
  13. 13 Sarvam 30B 2
  14. 14 kisanslm-gguf (Qwen3.5-2B base, Q4_K_M) 1
  15. 15 navarasa-2.0-7b (Gemma-7B base) 1
  16. 16 Qwen/Qwen2.5-7B-Instruct 1
  17. 17 qwen25vl-7b-base 1
  18. 18 param-1-2.9b 1
  19. 19 sarvam-1-2b 1
  20. 20 openhathi-7b 1
  21. 21 airavata-7b (Airavata-8bit, OpenHathi-7B base) 1

Source: Official Indic-KCC-Agri-Advisory-Benchmark leaderboard

#EntryScoreSettingDateSource
1 DeepSeek-V4-Flash 4 Apr 2026 Official leaderboard
2 Gemma 3 27B 3 Mar 2025 Official leaderboard
3 Qwen3.6 27B 3 Apr 2026 Official leaderboard
4 mistral-small-3.1-24b (anomaly flagged) 24B 3 Official leaderboard
5 gemma3-12b-int4 (QAT) 12B 2 Official leaderboard
6 Gemma 3 12B 2 Mar 2025 Official leaderboard
7 Phi-4 2 Dec 2024 Official leaderboard
8 gpt-oss-20b 2 Aug 2025 Official leaderboard
9 llama3.1-8b-instruct 8B 2 Official leaderboard
10 Qwen3-VL-8B-Instruct 2 Oct 2025 Official leaderboard
11 Gemma 3 4B 2 Mar 2025 Official leaderboard
12 krutrim-1-7b 7B 2 Official leaderboard
13 Sarvam 30B 2 Mar 2026 Official leaderboard
14 kisanslm-gguf (Qwen3.5-2B base, Q4_K_M) 2B 1 Official leaderboard
15 navarasa-2.0-7b (Gemma-7B base) 7B 1 Official leaderboard
16 Qwen/Qwen2.5-7B-Instruct 7.6B 1 Official leaderboard
17 qwen25vl-7b-base 7B 1 Official leaderboard
18 param-1-2.9b 2.9B 1 Official leaderboard
19 sarvam-1-2b 2B 1 Official leaderboard
20 openhathi-7b 7B 1 Official leaderboard
21 airavata-7b (Airavata-8bit, OpenHathi-7B base) 7B 1 Official leaderboard
#ModelScoreSettingDateSource
1 DeepSeek-V4-Flash 4 Apr 2026 Official leaderboard
2 Gemma 3 27B 3 Mar 2025 Official leaderboard
3 Qwen3.6 27B 3 Apr 2026 Official leaderboard
4 Gemma 3 12B 2 Mar 2025 Official leaderboard
5 Phi-4 2 Dec 2024 Official leaderboard
6 gpt-oss-20b 2 Aug 2025 Official leaderboard
7 Qwen3-VL-8B-Instruct 2 Oct 2025 Official leaderboard
8 Gemma 3 4B 2 Mar 2025 Official leaderboard
9 Sarvam 30B 2 Mar 2026 Official leaderboard

Rankings by part

We read 2 parts from the benchmark’s own leaderboard: English and Indic average. Each ranks models on one part of the test. DeepSeek-V4-Flash leads both. Arrows show how far a model moved from its overall rank among the models we track.

English

9 models we track have a score here.

  1. 1 DeepSeek-V4-Flash 4
  2. 2 Qwen3.6 27B ↑1 4
  3. 3 Gemma 3 27B ↓1 4
  4. 4 Qwen3-VL-8B-Instruct ↑3 3
  5. 5 Phi-4 3
  6. 6 Gemma 3 12B ↓2 3
  7. 7 gpt-oss-20b ↓1 2
  8. 8 Sarvam 30B ↑1 2
  9. 9 Gemma 3 4B ↓1 1

Number one over time

Lab claims count from the day the card came out, independent runs from the model’s release.

DeepSeek
Own leaderboard Not tested since Sep 2026

Source: Official Indic-KCC-Agri-Advisory-Benchmark leaderboard

Scores over time

All the scores we have, one dot each, so a model tested at several settings appears more than once.

Plot with other benchmarks
On the benchmark’s own leaderboard
1 2 3 4 5 2025 2026 1 2 3 4 5 2025 2026

Source: Official Indic-KCC-Agri-Advisory-Benchmark leaderboard

No lab model card reports it yet

We read the headline results of 540 launch posts from 57 labs, and none of them list this benchmark.

More in Multilingual

All 545 in Multilingual
  • MMMLU In use MMLU professionally translated into 14 languages. Multiple-choice accuracy. Claude Mythos Preview 92.7%
  • WMT24++ In use English-to-55-language translation of WMT24 test data. Scored with automatic metrics such as MetricX or xCOMET. K-EXAONE 90.5%
  • MMLU-ProX In use MMLU-Pro translated into up to 29 languages with expert review. Multiple-choice accuracy per language. Qwen3.7-Max 87%
  • CLIcK In use Models answer 1,995 multiple-choice questions on Korean language and culture. Questions come from official Korean exams and textbooks in eleven categories. Solar Open 2 90.7%
  • Global MMLU Lite 3.0 In use Models answer 400 human-translated MMLU questions per language in 23 languages. Version 3.0 adds Czech, Hungarian, Oriya, Slovak and Tajik and updates Italian. Accuracy. Claude Opus 5 95.7%
  • MMMB In use Models answer multiple-choice questions about images in six languages across 15 categories. The set has about 12,000 questions. LFM2.5-VL-3B 83%

Data audit

8 of 9 checks pass, 1 warning. Look first at the Newest score check. The newest came 7 months ago, in April 2026, for DeepSeek-V4-Flash. Every benchmark

Catalog record 2 pass
  • Pass Task access All tasks are public.
  • Pass Confidence Checked against a primary source.
Scores 1 warning 2 pass 2 notes
  • Pass Independent runs Scores from the official leaderboard (21 entries).
  • Pass Models scored 9 models we track, plus 12 other entries on its leaderboard.
  • Warning Newest score The newest came 7 months ago, in April 2026, for DeepSeek-V4-Flash.
  • Note Frontier models None of the top 10 has a score. The strongest tested is DeepSeek-V4-Flash (#87).
  • Note Headroom Scored as a rating, with no fixed ceiling.
Leaderboard import 3 pass
  • Pass Importer We read the board on github.com: 9 rows for models we track and 12 other entries at the last read.
  • Pass Last read November 20th, 2026.
  • Pass Read worked The last read worked.
Consistency 1 pass 1 note
  • Note Lab claims No model has both a lab number and an independent run.
  • Pass Known issues No published audit found problems.

How we check