YFarmX logoYFarmXIntelligence for an accelerating worldBriefings
YFarmX logoYFarmX
YFarmX

AI Desk · AI Fingerprinting · Compare

Meta: Llama 3.3 70B Instruct · Union Alpha

Identity similarity: highNear-exact agreement · 46/47Agreement this close usually means one tokenizer behind two serving templates, or a small vocabulary revision inside one family. The differing rows below say which strings part.

Meta: Llama 3.3 70B Instruct recordUnion Alpha recordOpen in the interactive bench

View all 50 test strings46 matched · 1 different · 3 excluded

String-by-string results

Expand a string to read or copy its exact text. Each row compares the two models’ token counts.

46 of 47 clean strings agree

Showing all 50 strings.

Test stringMeta: Llama 3.3 70B InstructUnion AlphaResult
English sentenceen-prose
Exact text
The quick brown fox jumps over the lazy dog near the riverbank.
14ExcludedExcluded
Long English sentenceen-long
Exact text
Attribution of a stealth model requires evidence that the model cannot choose to withhold.
16ExcludedExcluded
20 spacesspaces-20
Exact text
a                    b
Show whitespace marks

· = space, ⇥ = tab, ↵ = line break

a····················b
33✓ Matched
60 spacesspaces-60
Exact text
a                                                            b
Show whitespace marks

· = space, ⇥ = tab, ↵ = line break

a····························································b
33✓ Matched
20 tabstabs-20
Exact text
a																				b
Show whitespace marks

· = space, ⇥ = tab, ↵ = line break

a⇥⇥⇥⇥⇥⇥⇥⇥⇥⇥⇥⇥⇥⇥⇥⇥⇥⇥⇥⇥b
33✓ Matched
20 line breaksnewlines-20
Exact text
a



















b
Show whitespace marks

· = space, ⇥ = tab, ↵ = line break

a↵
↵
↵
↵
↵
↵
↵
↵
↵
↵
↵
↵
↵
↵
↵
↵
↵
↵
↵
↵
b
44✓ Matched
Mixed whitespacemixed-ws
Exact text
a 	
  	
   b
Show whitespace marks

· = space, ⇥ = tab, ↵ = line break

a·⇥↵
··⇥↵
···b
55✓ Matched
9 digitsdigits-9
Exact text
123456789
33✓ Matched
12 digitsdigits-12
Exact text
123456789012
44✓ Matched
30 repeated digitsdigits-30
Exact text
111111111111111111111111111111
1012≠ Different
Formatted numberdigits-sep
Exact text
1,234,567.89
77✓ Matched
Long decimalfloat-long
Exact text
3.14159265358979323846
99✓ Matched
Chinese sentencezh-common
Exact text
这是一个用于测试分词器行为的中文句子
1212✓ Matched
Long Chinese sentencezh-long
Exact text
人工智能模型的归属需要可验证的证据而不是自我报告
1818✓ Matched
Rare Chinese characterszh-rare
Exact text
龘齉齾爩鱻麤靐飝鞻
1818✓ Matched
Japanese kanaja-kana
Exact text
これはトークナイザーのテストです
1111✓ Matched
Japanese kanjija-kanji
Exact text
日本語の漢字と仮名の混在文章
12ExcludedExcluded
Koreanko
Exact text
한국어 토크나이저 테스트 문장입니다
1414✓ Matched
Russianru
Exact text
Проверка токенизатора на кириллице
1414✓ Matched
Arabicar
Exact text
اختبار المحلل اللغوي للنصوص العربية
1212✓ Matched
Hebrewhe
Exact text
בדיקת מנתח אסימונים בעברית
2424✓ Matched
Hindihi
Exact text
यह टोकनाइज़र परीक्षण वाक्य है
1515✓ Matched
Thaith
Exact text
นี่คือประโยคทดสอบตัวแบ่งคำ
1414✓ Matched
Greekel
Exact text
Δοκιμή του τμηματοποιητή κειμένου
1313✓ Matched
Basic emojiemoji-basic
Exact text
😀😃😄😁😆
1010✓ Matched
Emoji skin tonesemoji-skin
Exact text
👍🏻👍🏼👍🏽👍🏾👍🏿
3030✓ Matched
Joined family emojiemoji-zwj-family
Exact text
👨‍👩‍👧‍👦
1515✓ Matched
Repeated family emojiemoji-zwj-x3
Exact text
👨‍👩‍👧‍👦👨‍👩‍👧‍👦👨‍👩‍👧‍👦
4545✓ Matched
Flag emojiemoji-flags
Exact text
🇬🇧🇺🇸🇨🇳🇯🇵
2424✓ Matched
Profession emojiemoji-prof
Exact text
👩🏽‍💻👨🏿‍🚀🧑‍🔬
2626✓ Matched
Mathematical symbolsmath
Exact text
∀x∈ℝ ∃y∈ℚ : x⊕y ≡ ∅ ∧ ∇²φ = 0
3131✓ Matched
Box-drawing charactersboxdraw
Exact text
┌─┬─┐│├┼┤└┴┘╔═╗╚╝
2323✓ Matched
Combining accentscombining
Exact text
áêĩōü
1111✓ Matched
Extended Chinese characterscjk-ext-b
Exact text
𠀀𠀁𠀂𪛖
1313✓ Matched
Styled Unicode letterssurrogates
Exact text
𝕳𝖊𝖑𝖑𝖔 𝓦𝓸𝓻𝓵𝓭 𝟙𝟚𝟛
4040✓ Matched
Layered combining markszalgo
Exact text
h̸̢̛e̵̛�݃l̷̨͐l̸̛͝o̶̧͝
3535✓ Matched
Mixed writing directionsrtl-mix
Exact text
hello שלום world مرحبا
88✓ Matched
Python codepy-code
Exact text
def f(x):
    return [i**2 for i in range(x) if i % 3 == 0]
Show whitespace marks

· = space, ⇥ = tab, ↵ = line break

def·f(x):↵
····return·[i**2·for·i·in·range(x)·if·i·%·3·==·0]
2525✓ Matched
Indented Pythonpy-indent
Exact text
if a:
        if b:
                if c:
                        return 1
Show whitespace marks

· = space, ⇥ = tab, ↵ = line break

if·a:↵
········if·b:↵
················if·c:↵
························return·1
1515✓ Matched
JSON objectjson
Exact text
{"model":"ox-alpha","context_length":1048576,"top_p":0.95}
2020✓ Matched
HTML markuphtml
Exact text
<div class="row"><span id="a">x</span></div>
1616✓ Matched
Regular expressionregex
Exact text
^(?:[a-z0-9!#$%&*+/=?^_`{|}~-]+)@(?:[a-z0-9](?:[a-z0-9-]*)?\.)+$
4444✓ Matched
Camel-case identifiercamel
Exact text
getUserAccountBalanceByIdentifier
55✓ Matched
Snake-case identifiersnake
Exact text
get_user_account_balance_by_identifier
66✓ Matched
Repeated long wordrare-word-x5
Exact text
antidisestablishmentarianism antidisestablishmentarianism antidisestablishmentarianism antidisestablishmentarianism antidisestablishmentarianism
3030✓ Matched
Repeated wordrepeat-tok
Exact text
lorem lorem lorem lorem lorem lorem lorem lorem lorem lorem lorem lorem lorem lorem lorem lorem lorem lorem lorem lorem
2121✓ Matched
Base64 textbase64
Exact text
SGVsbG8gV29ybGQhIFRoaXMgaXMgYSB0ZXN0IHN0cmluZy4=
3535✓ Matched
Hexadecimal texthex
Exact text
deadbeefcafebabe0123456789abcdef
1111✓ Matched
URL texturl
Exact text
https://openrouter.ai/api/v1/models?limit=100&offset=0
1717✓ Matched
UUIDuuid
Exact text
3f2504e0-4f89-11d3-9a0c-0305e82c3301
2727✓ Matched

Counts show each string’s prompt-token cost after the fixed overhead is subtracted. Excluded readings are kept separate from matched and different results.

Published model settings

Limits, prices and capabilities

Compare the essentials, then open the developer settings for more detail.

AMeta: Llama 3.3 70B Instruct
BUnion Alpha

Context window

Different

Total token capacity per request

A131,072 tokens
B262,144 tokens

Maximum reply

Different

Published output limit in tokens

A16,384 tokens
B131,072 tokens

Input price

Different

US dollars per 1 million input tokens

AUS$0.22 / 1M tokens
BUS$0.00 / 1M tokens

Output price

Different

US dollars per 1 million output tokens

AUS$0.50 / 1M tokens
BUS$0.00 / 1M tokens

What you can send and receive

Different

ALlama 3.3 70B Instruct

Accepts
text
Produces
text

BUnion Alpha

Accepts
textimage
Produces
text

Developer settings

Open a section to compare the complete declarations.

Supported API settingsShared controls and the settings unique to each model6 shared · 11 model-specific

AOnly Llama 3.3 70B Instruct 11

  • Frequency penaltyfrequency_penalty
  • Token biaslogit_bias
  • Token probabilitieslogprobs
  • Minimum probabilitymin_p
  • Presence penaltypresence_penalty
  • Repetition penaltyrepetition_penalty
  • Seedseed
  • Stop sequencesstop
  • Structured outputsstructured_outputs
  • Top ktop_k
  • Top token probabilitiestop_logprobs

BOnly Union Alpha 0

All declared settings are shared.

6 settings supported by both models
  • Maximum tokensmax_tokens
  • Response formatresponse_format
  • Temperaturetemperature
  • Tool choicetool_choice
  • Toolstools
  • Top ptop_p
Reasoning settingsEffort levels and reasoning behaviourDeclarations available below
AUnspecifiedBUnspecified

Both models’ declarations are shown above.

Default settingsValues used when a request leaves a setting openDeclarations available below
ANo values declaredBNo values declared

Both models’ declarations are shown above.

Published API declarations. Prices and limits can vary by provider.

Method: tokenizer fingerprinting, v1, 50 strings. This page is static and dated; the interactive bench compares any two of the 353 measured models.