{"id":3648,"date":"2026-08-18T10:00:00","date_gmt":"2026-08-18T10:00:00","guid":{"rendered":"https:\/\/oddspapi.io\/blog\/?p=3648"},"modified":"2026-08-29T15:03:05","modified_gmt":"2026-08-29T15:03:05","slug":"sbobet-api-odds-access","status":"publish","type":"post","link":"https:\/\/oddspapi.io\/blog\/sbobet-api-odds-access\/","title":{"rendered":"SBOBet API: 10.5% on the Three-Way, 3.2% on the Handicap"},"content":{"rendered":"

SBOBet does not publish an API. There is no developer portal, no key, no docs page. The route people describe on forums runs through an agent account, and an agent account is a credit line with a person attached, not a data feed you can point a script at.<\/p>\n

You can still read SBOBet’s prices in Python. This post shows how, and then shows what the numbers say about the book once you have them, because SBOBet prices very differently from the US retail books sitting next to it in the same payload.<\/p>\n

One number frames the whole post. On a Brasileiro Serie A fixture two days out, SBOBet quoted 14 markets<\/strong>. DraftKings quoted 96. BetParx quoted 140. SBOBet was still the tightest sportsbook on the handicap.<\/p>\n

Why there is no SBOBet API<\/h2>\n

SBOBet runs an Asian-style operation. It takes large stakes on a narrow menu and it does not chase recreational volume through affiliate widgets or public data partnerships. A US book publishes odds feeds because odds feeds send traffic back. SBOBet’s business does not work that way, so the feed never got built.<\/p>\n

That leaves three options, and two of them are bad.<\/p>\n

\n\n\n\n\n\n\n\n
Route<\/th>\nWhat it costs you<\/th>\nWhat you get<\/th>\n<\/tr>\n<\/thead>\n
Agent account<\/td>\nCredit terms, a human intermediary, settlement risk<\/td>\nPrices you can bet, no machine-readable feed<\/td>\n<\/tr>\n
Scraping<\/td>\nSession handling, Cloudflare, per-region markup, ongoing breakage<\/td>\nOne book, brittle<\/td>\n<\/tr>\n
OddsPapi<\/td>\nA query parameter<\/td>\nSBOBet plus 350+ other books in one JSON response, with free historical price history<\/td>\n<\/tr>\n<\/tbody>\n<\/table>\n<\/figure>\n

OddsPapi carries SBOBet under the slug sbobet<\/code>. Same call, same shape, same parse as Pinnacle or Bet365. Everything below runs on the free tier.<\/p>\n

What SBOBet actually covers<\/h2>\n

Soccer is the core. Sampling one fixture per competition across the top of the board on 7 August 2026, SBOBet priced 12 of 14: LaLiga, Eredivisie, Liga Portugal, 2. Bundesliga, Ligue 2, EFL Cup, Chinese Super League, Brasileiro Serie B, Liga Profesional, Leagues Cup, Austrian Bundesliga, Armenian Premier League. It skipped J.League and a UEFA Conference League tie that only three books had opened.<\/p>\n

Down in the tail it thins out. Across 30 lower-division and amateur competitions sampled the same day, SBOBet appeared on 11. Pinnacle appeared on 20. If you assumed the Asian sharp goes deeper than Pinnacle into obscure leagues, the data says the opposite.<\/p>\n

Other sports, five fixtures sampled each:<\/p>\n

\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n
Sport<\/th>\nsportId<\/th>\nFixtures with odds (7 days)<\/th>\nSBOBet present<\/th>\n<\/tr>\n<\/thead>\n
Tennis<\/td>\n12<\/td>\n200<\/td>\n4 of 5<\/td>\n<\/tr>\n
Handball<\/td>\n22<\/td>\n18<\/td>\n4 of 5<\/td>\n<\/tr>\n
Ice Hockey<\/td>\n15<\/td>\n15<\/td>\n3 of 4<\/td>\n<\/tr>\n
Volleyball<\/td>\n23<\/td>\n4<\/td>\n3 of 4<\/td>\n<\/tr>\n
Basketball<\/td>\n11<\/td>\n67<\/td>\n2 of 5<\/td>\n<\/tr>\n
American Football<\/td>\n14<\/td>\n8<\/td>\n1 of 5<\/td>\n<\/tr>\n
Baseball<\/td>\n13<\/td>\n154<\/td>\n0 of 5<\/td>\n<\/tr>\n
Cricket<\/td>\n27<\/td>\n100<\/td>\n0 of 5<\/td>\n<\/tr>\n
MMA<\/td>\n20<\/td>\n32<\/td>\n0 of 5<\/td>\n<\/tr>\n
CS2<\/td>\n17<\/td>\n46<\/td>\n0 of 5<\/td>\n<\/tr>\n<\/tbody>\n<\/table>\n<\/figure>\n

Write your MLB and UFC code against other books. The catalogue also flags sbobet<\/code> with liveOdds: false<\/code>, so treat this slug as a pre-match source and build in-play features on books that carry them.<\/p>\n

The 14-market menu<\/h2>\n

Here is the entire SBOBet menu on Botafogo v Fluminense (id1000032566886900<\/code>), pulled 7 August 2026 with kick-off two days out:<\/p>\n

\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n
Market ID<\/th>\nName<\/th>\nPrices<\/th>\n<\/tr>\n<\/thead>\n
101<\/td>\nFull Time Result<\/td>\n2.33 \/ 3.18 \/ 2.79<\/td>\n<\/tr>\n
1072<\/td>\nAsian Handicap 0<\/td>\n1.775 \/ 2.12<\/td>\n<\/tr>\n
1070<\/td>\nAsian Handicap -0.25<\/td>\n2.14 \/ 1.765<\/td>\n<\/tr>\n
1010<\/td>\nOver Under 2.5<\/td>\n1.995 \/ 1.875<\/td>\n<\/tr>\n
10170<\/td>\nOver Under 2.25<\/td>\n1.705 \/ 2.19<\/td>\n<\/tr>\n
10208<\/td>\nFirst Half Result<\/td>\n2.99 \/ 2.13 \/ 3.15<\/td>\n<\/tr>\n
10216<\/td>\nFirst Goal<\/td>\n1.86 \/ 2.06 \/ 10.0<\/td>\n<\/tr>\n
10219<\/td>\nLast Goal<\/td>\n1.86 \/ 2.06 \/ 10.0<\/td>\n<\/tr>\n
10222<\/td>\nOdd \/ Even<\/td>\n1.915 \/ 1.975<\/td>\n<\/tr>\n
10336<\/td>\nCorrect Score<\/td>\n25 outcomes<\/td>\n<\/tr>\n
101902<\/td>\nDouble Chance<\/td>\n1.37 \/ 1.52 \/ 1.29<\/td>\n<\/tr>\n
101919<\/td>\nHalf Time \/ Full Time<\/td>\n9 outcomes<\/td>\n<\/tr>\n
102462<\/td>\nCorrect Score 1st Half<\/td>\n16 outcomes<\/td>\n<\/tr>\n
102526<\/td>\nCorrect Score 2nd Half<\/td>\n16 outcomes<\/td>\n<\/tr>\n<\/tbody>\n<\/table>\n<\/figure>\n

That same list came back on all five fixtures I pulled, across five competitions on three continents. No both-teams-to-score. No corners. No team totals. No player props. No alt ladder. Every other book on that fixture quoted BTTS; SBOBet quoted zero corner markets while BetParx quoted 54.<\/p>\n

The handicaps and totals are the tell. SBOBet posted two handicap rungs and two goal lines. Pinnacle posted nine handicap rungs and eight goal lines. SBOBet publishes its estimate of the true number and takes bets on either side of it. The menu is the opinion.<\/p>\n

Step 1: authenticate and confirm the slug<\/h2>\n

The API key rides as a query parameter. It is never a header.<\/p>\n

import requests, time\n\nAPI_KEY  = \"YOUR_API_KEY\"\nBASE_URL = \"https:\/\/api.oddspapi.io\/v4\"\n\ndef api(path, **params):\n    params[\"apiKey\"] = API_KEY\n    for _ in range(5):\n        r = requests.get(f\"{BASE_URL}\/{path}\", params=params, timeout=120)\n        if r.status_code == 200:\n            return r.json()\n        retry = r.json().get(\"error\", {}).get(\"retryMs\", 1500) \/ 1000\n        time.sleep(retry + 0.3)\n    r.raise_for_status()\n\nbooks = api(\"bookmakers\")\nprint(len(books), \"bookmakers\")\nprint([b for b in books if b[\"slug\"] == \"sbobet\"])<\/code><\/pre>\n
348 bookmakers\n[{'bookmakerName': 'SBOBET', 'slug': 'sbobet', 'liveOdds': False, 'cloneOf': None}]<\/code><\/pre>\n
The retry block matters. The free tier rate-limits per endpoint and returns HTTP 429 with a retryMs<\/code> value in the body. That body is valid JSON, so code that checks for a bookmakerOdds<\/code> key without checking the status code reads a rate-limit response as “no coverage” and reports that SBOBet prices nothing.<\/p>\n
Step 2: find a fixture SBOBet is pricing<\/h2>\n
\/fixtures<\/code> takes a sport and a date range up to 10 days wide. Only fixtures with hasOdds: true<\/code> return a price payload.<\/p>\n
fixtures = api(\"fixtures\", sportId=10, **{\"from\": \"2026-08-08\", \"to\": \"2026-08-10\"})\nlive = [f for f in fixtures if f[\"hasOdds\"]]\n\nfor f in live[:5]:\n    print(f[\"fixtureId\"], f[\"tournamentName\"],\n          f[\"participant1Name\"], \"v\", f[\"participant2Name\"])<\/code><\/pre>\n
Team names live on participant1Name<\/code> and participant2Name<\/code>. Do not query \/odds<\/code> with the bookmakers<\/code> filter set to sbobet<\/code> until you know SBOBet is on the fixture. That filter zeroes the entire response if any requested slug is missing, which looks identical to a book with no prices.<\/p>\n
Step 3: parse the menu<\/h2>\n
Market IDs are integers and the names live in \/markets<\/code>. Build the lookup once and reuse it. The catalogue is global rather than per-sport, so sportId<\/code> on that endpoint changes nothing; read the market IDs off the live payload and use the catalogue to name them.<\/p>\n
catalog = api(\"markets\", sportId=10)\nMARKET  = {m[\"marketId\"]: m for m in catalog}\nOUTCOME = {(m[\"marketId\"], o[\"outcomeId\"]): o[\"outcomeName\"]\n           for m in catalog for o in m.get(\"outcomes\", [])}\n\ndef quotes(book, market_id):\n    \"\"\"Live two-sided prices for one market on one book.\"\"\"\n    market = book[\"markets\"].get(str(market_id))\n    if not market:\n        return {}\n    live = {}\n    for outcome_id, outcome in market[\"outcomes\"].items():\n        price = outcome[\"players\"].get(\"0\")\n        if price and price.get(\"active\") is not False and price.get(\"price\", 0) > 1:\n            live[int(outcome_id)] = price[\"price\"]\n    return live\n\nFIXTURE = \"id1000032566886900\"\nboard   = api(\"odds\", fixtureId=FIXTURE)[\"bookmakerOdds\"]\nsbobet  = board[\"sbobet\"]\n\nprint({slug: len(b[\"markets\"]) for slug, b in sorted(board.items())})\n\nfor market_id in sorted(sbobet[\"markets\"], key=int):\n    mid  = int(market_id)\n    meta = MARKET.get(mid, {})\n    name = meta.get(\"marketName\", \"?\")\n    if meta.get(\"handicap\"):\n        name += f\" {meta['handicap']}\"\n    priced = quotes(sbobet, mid)\n    if len(priced) <= 4:\n        print(f\"{mid:>7}  {name:34s} \" +\n              \"  \".join(f\"{OUTCOME.get((mid, o), o)} {p}\" for o, p in priced.items()))\n    else:\n        print(f\"{mid:>7}  {name:34s} ({len(priced)} outcomes)\")<\/code><\/pre>\n
{'ballybet': 131, 'bet365': 56, 'betmgm': 72, 'betparx': 140, 'betrivers': 131,\n 'borgata': 72, 'caesars': 57, 'draftkings': 96, 'fanduel': 110, 'fourwinds': 75,\n 'hardrockbet': 94, 'kalshi': 27, 'pinnacle': 88, 'pointsbet.com.au': 33,\n 'polymarket': 60, 'sbobet': 14, 'williamhill': 57}\n\n    101  Full Time Result                   2 2.79  X 3.18  1 2.33\n   1010  Over Under Full Time 2.5           Under 1.875  Over 1.995\n   1070  Asian Handicap -0.25               2 1.765  1 2.14\n   1072  Asian Handicap                     1 1.775  2 2.12\n  10170  Over Under Full Time 2.25          Under 2.19  Over 1.705\n  10208  First Half Result                  1 2.99  X 2.13  2 3.15\n  ...<\/code><\/pre>\n
The players<\/code> dict is keyed \"0\"<\/code> on game lines. On player-prop markets it is keyed by player ID instead, which is why hardcoding players[\"0\"]<\/code> makes prop markets look empty. SBOBet ships no props, so \"0\"<\/code> is safe for this book.<\/p>\n
Step 4: match SBOBet’s line at other books<\/h2>\n
This is the part that breaks naive line shopping. Every handicap and every goal line has its own market ID<\/strong>. Over\/Under 2.5 is market 1010, Over\/Under 2.25 is 10170, Over\/Under 2.75 is 10172. If you hardcode 1010 and call it “the total”, SBOBet vanishes on most fixtures, because it often does not quote 2.5 at all.<\/p>\n
Across the five fixtures I sampled, SBOBet’s goal lines were 2.25 alone, then 2.25 and 2.5, then 2.5 and 2.75, then 3 and 3.25, then 2.75 and 3. Two of five included 2.5. On the Eredivisie fixture SBOBet sat on 3 and 3.25 while Caesars, William Hill, BetMGM and FanDuel all quoted 2.5 and nothing near 3.25.<\/p>\n
Resolve by name and handicap, then compare like for like.<\/p>\n
def lines(book, market_name, period=\"fulltime\"):\n    \"\"\"{handicap: (market_id, prices)} for one market name on one book.\"\"\"\n    found = {}\n    for market_id in book[\"markets\"]:\n        meta = MARKET.get(int(market_id))\n        if meta and meta[\"marketName\"] == market_name and meta.get(\"period\") == period:\n            priced = quotes(book, market_id)\n            if len(priced) == 2:\n                found[meta[\"handicap\"]] = (int(market_id), priced)\n    return found\n\ndef margin(prices):\n    return sum(1 \/ p for p in prices.values()) - 1\n\nfor line, (mid, priced) in sorted(lines(sbobet, \"Over Under Full Time\").items()):\n    print(f\"\\nline {line} (market {mid})\")\n    for slug, book in sorted(board.items()):\n        other = lines(book, \"Over Under Full Time\").get(line)\n        if other:\n            print(f\"  {slug:18s} {other[1]}  margin {margin(other[1])*100:5.2f}%\")<\/code><\/pre>\n
line 2.25 (market 10170)\n  ballybet           {10170: 1.72, 10171: 2.1}    margin  5.76%\n  betparx            {10170: 1.72, 10171: 2.1}    margin  5.76%\n  betrivers          {10171: 2.1, 10170: 1.72}    margin  5.76%\n  draftkings         {10170: 1.667, 10171: 2.05}  margin  8.77%\n  pinnacle           {10170: 1.709, 10171: 2.19}  margin  4.18%\n  sbobet             {10171: 2.19, 10170: 1.705}  margin  4.31%\n\nline 2.5 (market 1010)\n  ...\n  fanduel            {1010: 2.02, 1011: 1.81}     margin  4.75%\n  pinnacle           {1011: 1.9, 1010: 1.97}      margin  3.39%\n  sbobet             {1011: 1.875, 1010: 1.995}   margin  3.46%\n  ...<\/code><\/pre>\n
Six of 17 books quoted the 2.25 line. Sixteen quoted 2.5. Quarter lines are an Asian-market convention and most US books skip them, so a cross-book comparison built on 2.5 alone throws away the line SBOBet and Pinnacle think is right.<\/p>\n
The margin split: wide on the three-way, tight on the handicap<\/h2>\n
Run the margins across both market types on the same fixture and SBOBet looks like two different books.<\/p>\n
for slug in (\"sbobet\", \"pinnacle\", \"draftkings\", \"bet365\", \"fanduel\"):\n    book = board.get(slug)\n    three = quotes(book, 101)\n    ah    = lines(book, \"Asian Handicap\")\n    best  = min((margin(p) for _, p in ah.values()), default=None)\n    tail  = f\"best AH {best*100:6.2f}% ({len(ah)} rungs)\" if best else \"no AH\"\n    print(f\"{slug:12s} 1X2 {margin(three)*100:6.2f}%   {tail}\")<\/code><\/pre>\n
sbobet       1X2  10.21%   best AH   3.39% (2 rungs)\npinnacle     1X2   4.42%   best AH   3.04% (9 rungs)\ndraftkings   1X2   6.74%   best AH   8.85% (7 rungs)\nbet365       1X2   7.15%   no AH\nfanduel      1X2   6.02%   best AH   7.63% (2 rungs)<\/code><\/pre>\n
SBOBet charges 10.21% on the three-way and 3.39% on its handicap. Its own margin drops by a factor of three depending on which market you click. DraftKings runs the opposite way: 6.74% on the three-way and 8.85% on the handicap.<\/p>\n
Widen the sample to five fixtures across the Champions League qualifiers, Brasileiro Serie A, the English Championship, the Eredivisie and Liga Portugal, and the pattern holds. Median margin per book:<\/p>\n
\n\n\n\n\n\n\n\n\n\n\n\n\n
Book<\/th>\n1X2<\/th>\nTightest handicap<\/th>\nTightest goal line<\/th>\n<\/tr>\n<\/thead>\n
SBOBet<\/td>\n10.50%<\/td>\n3.17%<\/strong><\/td>\n3.46%<\/td>\n<\/tr>\n
Pinnacle<\/td>\n6.14%<\/td>\n4.42%<\/td>\n4.92%<\/td>\n<\/tr>\n
Hard Rock Bet<\/td>\n7.30%<\/td>\n5.46%<\/td>\n6.66%<\/td>\n<\/tr>\n
Caesars \/ William Hill<\/td>\n7.92%<\/td>\n6.82%<\/td>\n5.98%<\/td>\n<\/tr>\n
FanDuel<\/td>\n8.10%<\/td>\n7.63%<\/td>\n4.89%<\/td>\n<\/tr>\n
BetMGM<\/td>\n5.91%<\/td>\n8.78%<\/td>\n4.72%<\/td>\n<\/tr>\n
DraftKings<\/td>\n6.91%<\/td>\n9.11%<\/td>\n7.50%<\/td>\n<\/tr>\n
Bet365<\/td>\n8.54%<\/td>\nquotes none<\/td>\n2.24%<\/td>\n<\/tr>\n<\/tbody>\n<\/table>\n<\/figure>\n
SBOBet posts the worst three-way price in the field and the best handicap price in the field. Benchmark your model against its handicap. Ignore its 1X2, or use it as a demonstration of what a book charges on a market it does not want. We ran the full version of this study across nine Premier League fixtures in Why Sharps Bet Asian Handicap<\/a>, and the ratios came out the same.<\/p>\n
De-vig the handicap<\/h2>\n
Two outcomes, one line, no draw. Multiply each price by the overround and you have SBOBet’s fair number.<\/p>\n
mid, priced = min(lines(sbobet, \"Asian Handicap\").values(), key=lambda v: margin(v[1]))\noverround   = 1 + margin(priced)\n\nprint(f\"market {mid} ({MARKET[mid]['handicap']}), margin {margin(priced)*100:.2f}%\")\nfor outcome_id, price in priced.items():\n    fair = price * overround\n    print(f\"  {OUTCOME[(mid, outcome_id)]}: {price}  ->  fair {fair:.3f}  ({100\/fair:.1f}%)\")<\/code><\/pre>\n
market 1070 (-0.25), margin 3.39%\n  2: 1.765  ->  fair 1.825  (54.8%)\n  1: 2.14   ->  fair 2.212  (45.2%)<\/code><\/pre>\n
Fluminense at 54.8% on a quarter-ball line, from a book with a 3.39% cut. Compare that to what you get de-vigging a 10% three-way and the reason Asian syndicates live on the handicap becomes obvious. For the settlement mechanics of quarter lines, see the Asian Handicap calculator<\/a>. For three-way de-vigging methods, see No-Vig Odds<\/a>.<\/p>\n
Free historical odds: where SBOBet’s attention goes<\/h2>\n
\/historical-odds<\/code> returns the full snapshot history on the free tier, up to three bookmakers per call. The response key is bookmakers<\/code>, not bookmakerOdds<\/code>, and players[\"0\"]<\/code> is a list rather than a dict.<\/p>\n
hist = api(\"historical-odds\", fixtureId=FIXTURE, bookmakers=\"sbobet,pinnacle\")\n\nfor slug, book in hist[\"bookmakers\"].items():\n    print(f\"\\n{slug}: {len(book['markets'])} markets with history\")\n    for market_id in (\"101\", \"1072\"):\n        for outcome_id, outcome in book[\"markets\"][market_id][\"outcomes\"].items():\n            snaps   = outcome[\"players\"][\"0\"]\n            changes = sum(1 for a, b in zip(snaps, snaps[1:]) if a[\"price\"] != b[\"price\"])\n            print(f\"  market {market_id} outcome {outcome_id}: \"\n                  f\"{len(snaps):>4} snapshots, {changes:>4} price changes, \"\n                  f\"{snaps[0]['price']} -> {snaps[-1]['price']}\")<\/code><\/pre>\n
pinnacle: 88 markets with history\n  market 101  outcome 101 :   33 snapshots,   23 price changes, 2.51  -> 2.52\n  market 1072 outcome 1072:   33 snapshots,   23 price changes, 1.793 -> 1.769\n\nsbobet: 14 markets with history\n  market 101  outcome 101 :   74 snapshots,    3 price changes, 2.44  -> 2.33\n  market 1072 outcome 1072:  284 snapshots,  226 price changes, 1.87  -> 1.775<\/code><\/pre>\n
Over five days SBOBet moved its three-way price three times and its handicap 226 times. Pinnacle repriced both at the same rate. Count price changes rather than snapshots, because the feed records on a cadence and most snapshots repeat the previous price.<\/p>\n
That 75-to-1 ratio is the cleanest evidence in the payload of where a book’s opinion lives. If you are building a steam detector or a closing-line study on SBOBet, watch market 1072 and ignore market 101. See the steam move detector<\/a> for the alerting side of that.<\/p>\n
Four things that will bite you<\/h2>\n